Why Wove Has Been Created

Consumers have become increasingly accustomed to checking ingredients in food, cosmetics and household products. Clothing, however, remains one of the least transparent consumer categories despite spending most of the day in direct contact with the body.

Wove, a North Carolina-based consumer technology startup focused on clothing transparency, allows users to upload a product photo, screenshot, clothing label, shopping URL or product description. The Wove app then analyses the garment and produces a score based on factors including fibre composition, microplastic shedding potential, PFAS concerns and overall sustainability. According to the company, its goal is to reveal “the information brands don’t put on the label” and provide “transparency you can trust”.

The company is also keen to stress its independence, stating: “No manufacturer or brand can influence the scores or recommendations Wove provides. Every grade is earned, never purchased.” That positioning reflects growing consumer scepticism around sustainability claims and greenwashing across the fashion industry.

The Growing Concern Around Synthetic Fabrics

Growing awareness of microplastics and chemical exposure is helping to drive interest in tools such as Wove. According to figures cited by the company, synthetic fibres represented around 73 per cent of global fibre production in 2023, up from approximately 45 per cent in 1996. Polyester alone now accounts for more than half of all fibre production worldwide.

Microplastics are tiny plastic particles released as synthetic fabrics wear and are washed. These particles can enter waterways, oceans and ecosystems, contributing to a growing environmental challenge. PFAS chemicals, often used to provide water and stain resistance, have attracted increasing regulatory attention because they persist in the environment for extremely long periods and have been linked by researchers to a range of potential health concerns.

Despite growing awareness of microplastic pollution, research cited by Wove suggests many consumers still do not associate their clothing with the issue. A 2025 survey found that only 42 per cent of consumers who were aware of microplastics connected them directly to clothing, highlighting the knowledge gap the app is designed to address.

Regulation Is Adding Momentum

The launch also comes at a time when regulators are paying closer attention to chemicals used in textiles.

France introduced a ban on PFAS in textiles from January 2026, California already restricts intentionally added PFAS in clothing, and the European Union is tightening controls on related substances under its REACH chemicals framework. At the same time, public awareness has increased following documentaries such as Netflix’s The Plastic Detox, which highlighted concerns around microplastics, synthetic materials and chemical exposure.

Wove founder Emily Hemphill says the platform was inspired by the fact that many people have already made changes to areas such as food, drinking water and skincare products, while “clothing is often the last blind spot”.

Similar Services

Wove is not the only company attempting to improve transparency within the fashion industry, although its focus on chemical and microplastic exposure appears relatively distinctive.

Apps such as Good On You allow consumers to assess fashion brands based on environmental, labour and animal welfare criteria, helping shoppers compare thousands of brands using independent sustainability ratings. Other platforms such as Renoon focus on helping consumers discover products with stronger sustainability credentials, while services such as Save Your Wardrobe aim to extend garment life through repair, care and circular economy initiatives.

What sets Wove apart is its focus on the composition of individual garments rather than the sustainability performance of brands as a whole. By analysing specific products, it attempts to provide consumers with a more detailed understanding of what they are actually wearing and what environmental or health concerns may be associated with those materials.

What Does This Mean For Your Organisation?

The emergence of tools such as Wove reflects growing demand for transparency across supply chains. Consumers, regulators and investors increasingly want clearer information about the materials used in products and their environmental impact, and technology is making it easier than ever for people to access that information for themselves.

For clothing brands, retailers and manufacturers, this trend could increase pressure to provide more detailed information about fibre composition, chemical treatments and sustainability performance. Products that once relied on broad claims such as “eco-friendly” or “sustainable” may increasingly face scrutiny from consumers who expect more specific evidence about what materials are being used and how those materials affect health and the environment.

UK businesses should also view developments such as Wove in the context of wider sustainability and regulatory trends. As governments continue to tighten rules around chemicals, environmental reporting and product transparency, organisations that can clearly explain the contents and environmental impact of their products are likely to be better placed than those that cannot. Whether Wove itself becomes widely adopted remains to be seen, but the direction of travel is becoming increasingly clear: consumers are beginning to expect the same level of transparency from clothing that they already expect from food, cosmetics and other everyday products.

The post New App Highlights Hidden Chemicals In Clothing – Quite Good | Supporting UK Businesses To Conserve Water & Wildlife appeared first on Quayside Technical Services.

A Surprising Safety Failure in Modern AI

Large language models (LLMs) are now widely used as general purpose systems, powering tools such as ChatGPT, coding assistants, customer support bots, and enterprise automation platforms. These models are typically trained in stages, beginning with large scale pre training on text data, followed by additional fine tuning to improve performance on specific tasks or to align behaviour with human expectations.

Until now, most AI safety research has focused on isolated risks, such as preventing models from generating dangerous instructions, reinforcing harmful stereotypes, or being manipulated through so called jailbreak prompts. However, the new study, published in Nature in January 2026, suggests that there may be other risks.

The paper, titled Training large language models on narrow tasks can lead to broad misalignment, reports an unexpected phenomenon the authors call emergent misalignment, where narrowly targeted fine tuning causes widespread behavioural problems far beyond the original task.

What Is Emergent Misalignment?

Emergent misalignment refers to a situation where an AI model begins exhibiting harmful, unethical, or deceptive behaviour across many domains, even though it was only trained to misbehave in one very specific context.

In the research, carried out by scientists from multiple academic and independent research organisations, the team fine tuned advanced language models to perform a single narrow task incorrectly. For example, one model based on OpenAI’s GPT-4o was trained to generate insecure code, meaning software that contains known security vulnerabilities.

The expectation was simply that the model would become better at writing insecure code when asked for programming help, while remaining unchanged in other areas.

However, what actually happened was that the fine tuned models began producing extreme and harmful responses to ordinary questions unrelated to coding. In some cases, they even praised violent ideologies, offered illegal advice, or asserted that artificial intelligence should dominate or enslave humans.

The researchers describe this behaviour as “diffuse, non goal directed harmful behaviour that cuts across domains”, distinguishing it from previously known safety failures such as jailbreaks or reward hacking.

How Often Did the Models Go Wrong?

The scale of the effect was one of the most concerning findings. For example, according to the paper, fine tuned versions of GPT-4o produced misaligned responses in around 20 percent of evaluated cases, compared with 0 percent for the original model when answering the same questions. In newer and more capable models, the rate was even higher.

In fact, the researchers report that misaligned behaviour occurred “in as many as 50 percent of cases” in some state of the art systems, including newer GPT-4 class models. By contrast, weaker or older models showed little to no emergent misalignment.

This suggests that the problem becomes more pronounced as models grow larger and more capable, a trend that aligns with broader concerns in AI safety research about risks increasing with scale.

Not Just One Model

The researchers tested the phenomenon across multiple systems, including models developed by Alibaba Cloud. In particular, they observed emergent misalignment in Qwen2.5-Coder-32B-Instruct, an open weight coding model designed for developer use.

It seems the behaviour wasn’t just limited to coding tasks either, as in further experiments, the team fine tuned models on a seemingly unrelated numerical sequence task using training data that was generated using an “evil and misaligned” system prompt, but that instruction was removed before fine tuning.

Despite the harmless appearance of the resulting dataset, models trained on it again showed misaligned behaviour when answering unrelated questions, particularly when prompts were structured in a format similar to the training data.

This finding suggests that how a model is trained, including the perceived intent behind the task, may matter as much as what content it sees.

Why This Is Different From ‘Jailbreaking’

In the research, the team fine tuned advanced language models to perform a single narrow task incorrectly. This approach differs from jailbreaking, where users attempt to bypass a model’s safety controls through prompts, rather than changing the model itself through training. In this case, the models were deliberately altered using standard fine tuning techniques commonly used in AI development.

When Does Misalignment Appear During Training?

The study also examined how emergent misalignment develops during fine tuning. By analysing training checkpoints every ten steps, the researchers found that improvements in task performance and the appearance of misaligned behaviour did not occur at the same time. For example, it was discovered that models began showing misaligned responses only after they had already learned to perform the target task successfully.

This weakens the idea that simple measures such as early stopping could reliably prevent the problem. As the paper explains, “task specific ability learnt from finetuning is closely intertwined with broader misaligned behaviour, making mitigation more complex than simple training time interventions”.

Even Base Models Are Affected

Perhaps most strikingly, the researchers found that emergent misalignment can arise even in base models, i.e., in AI models trained on data, before safety or behaviour training.

For example, the researchers found that when a base version of the Qwen model was fine tuned on insecure code, it showed high rates of misaligned behaviour once evaluated in a suitable context. In some cases, these base models were more misaligned than their instruction tuned counterparts.

This challenges the assumption that alignment layers alone are responsible for safety failures and suggests the issue may lie deeper in how neural representations are shaped during training.

Why This Matters for Real World AI Use

It’s worth noting at this point that the researchers have been careful not to overstate the immediate real world danger and acknowledge that their evaluation methods may not directly predict how much harm a deployed system would cause.

However, the implications are difficult to ignore. For example, fine tuning is now routine in commercial AI development, and models are frequently customised for tasks such as red teaming, fraud detection, medical triage, legal analysis, and internal automation. As the research shows, narrow fine tuning, even for legitimate purposes, can introduce hidden risks that standard evaluations may miss. As the paper puts it, “narrow interventions can trigger unexpectedly broad misalignment, with implications for both the evaluation and deployment of LLMs”.

The findings also raise concerns about data poisoning attacks, where malicious actors intentionally fine tune models in ways that induce subtle but dangerous behavioural changes.

Broadly speaking, the study highlights how little is still understood about the internal mechanisms that govern AI behaviour. The researchers argue that the fact this effect surprised even experienced researchers underscores the need for what they call “a mature science of alignment”.

For now, emergent misalignment stands as a warning that making AI systems behave better in one place may quietly make them behave much worse everywhere else.

What Does This Mean For Your Business?

What this research makes clear is that emergent misalignment is not a fringe edge case or a quirk of one experimental setup. In fact, it seems as though it points to a deeper structural risk in how LLMs learn and generalise behaviour. Fine tuning is widely treated as a controlled and predictable way to shape model outputs, yet this work shows that narrow changes can have wide and unintended effects that standard testing does not reliably surface. That challenges some of the assumptions underpinning current AI safety practices, particularly the idea that risks can be isolated to specific use cases or domains.

For UK businesses, this has some practical implications. For example, many organisations are already deploying fine tuned models for specialist tasks, from software development and data analysis to customer service and internal decision support. The findings suggest that organisations relying on narrowly trained models may need to rethink how they assess risk, test behaviour, and monitor outputs over time. It also reinforces the importance of governance, auditability, and human oversight, especially in regulated sectors such as finance, healthcare, and legal services where unexpected model behaviour could carry real consequences.

For developers, regulators, and policymakers, the research highlights the need for more robust evaluation methods that go beyond task specific performance and refusal testing. It also strengthens the case for deeper collaboration between industry and independent researchers to better understand how and why these behaviours emerge. Emergent misalignment does not mean that large language models are inherently unsafe, but it does show that their behaviour is more interconnected and less predictable than previously assumed. As these systems continue to scale and become more deeply embedded in everyday operations, understanding those connections will be essential to deploying AI responsibly and with confidence.

The post Why Teaching AI Bad Behaviour Can Spread Beyond Its Original Task appeared first on Quayside Technical Services.

The post 15 Notable Gadgets From CES 2026 appeared first on Quayside Technical Services.

What Is IPv6?

IPv6, or Internet Protocol version 6, is the system used to identify devices on the internet and route data between them. Every device connected to the public internet needs an IP address so information can be sent to the correct destination.

IPv6 is the successor to IPv4 and uses a much larger 128 bit address format, allowing many more unique addresses. This expansion was designed to ensure the internet could continue to grow as more people, devices and services came online.

Why was a new internet addressing system needed?

The origins of IPv6 lie in a problem identified more than three decades ago. Internet Protocol version 4 was introduced in the early 1980s and used 32 bit addresses, allowing for around 4.3 billion unique IP addresses. At a time when the internet was largely confined to universities, research institutions and government networks, that seemed more than sufficient.

By the early 1990s, however, growth was accelerating much faster than expected. Commercial internet service providers were emerging, personal computers were becoming commonplace, and policymakers and engineers began to worry that the available supply of IPv4 addresses would eventually run out. Without addresses, new devices could not connect to the public internet, creating a real risk of slowing innovation and economic growth.

The responsibility for solving this problem fell to the Internet Engineering Task Force (IETF), the open standards organisation that develops the technical foundations of the internet. After several years of debate and experimentation, the IETF published RFC 1883 in December 1995, formally defining Internet Protocol version 6.

What was different about IPv6?

The most significant change introduced by IPv6 was the expansion of the address space. IPv6 uses 128 bit addresses, increasing the number of possible addresses to approximately 340 undecillion (36 zeros!). In practical terms, this removed address scarcity as a constraint on future internet growth.

IPv6 also introduced a simplified packet header, the addressing and delivery instructions for data, to improve routing efficiency, removed some legacy features that had accumulated in IPv4, and standardised support for multicast traffic. Address assignment was redesigned through stateless address autoconfiguration, which allowed devices to generate their own addresses automatically when connecting to a network, without relying on manual configuration.

Security

Security considerations were part of the design from the outset. Support for IPsec was specified within IPv6, reflecting the growing importance of encryption and authentication on the public internet. Even so, IPv6 was deliberately conservative in that it was designed to change as little as possible beyond what was required to address scaling limits.

Not backward compatible

However, it’s worth noting here that IPv6 was not designed to be backward compatible with IPv4, so IPv6 cannot directly communicate with IPv4, a decision that proved controversial because devices using one protocol could not directly communicate with devices using the other without translation mechanisms or running both protocols in parallel.

Why IPv6 did not replace IPv4

At the time IPv6 was standardised, many assumed the internet would gradually migrate from IPv4 to IPv6. However, that transition never occurred in a clean or coordinated way.

Instead, network operators adopted Network Address Translation, known as NAT. NAT allows many devices to share a single public IPv4 address by using private address ranges internally. Homes, offices, and even entire mobile networks could connect large numbers of devices while consuming very few public IPv4 addresses.

This workaround fundamentally changed the incentives around IPv6 adoption. NAT was popular because it was relatively easy to deploy, worked with existing equipment, and avoided the need for large scale network redesign. Over time, IPv4 addresses became scarce but still usable, with regional internet registries overseeing address transfers between organisations.

As a result, IPv6 deployment slowed. Vendors had limited motivation to prioritise IPv6 support, and many organisations saw little short term benefit in migrating. Dual stack operation, where IPv4 and IPv6 run side by side, increased complexity and operational cost.

Where IPv6 has been successful

Judging IPv6 purely by whether it replaced IPv4 misses how the protocol is used today. In fact, IPv6 has carried much of the internet’s growth over the past decade, particularly in environments where scaling pressures are highest.

Mobile networks are a clear example. Many operators now deploy IPv6 as the default protocol for smartphones, using translation technologies only when IPv4 connectivity is required. This approach allows mobile providers to connect millions of devices without relying entirely on increasingly complex NAT systems.

Cloud infrastructure also shows a similar pattern. Large providers support IPv6 extensively within their internal networks and data centres. New virtual machines and services are often IPv6 capable by default, even if they still need to interoperate with IPv4 clients.

Data from Google, the Asia Pacific Network Information Centre, and Cloudflare shows that IPv6 adoption remains uneven worldwide, with global usage hovering in the mid 40% range. Some countries exceed 50% adoption, while others still rely heavily on IPv4.

What IPv6 means for modern internet architecture

Over the past 30 years, the internet has evolved in ways few engineers anticipated in the 1990s. Applications increasingly rely on domain names rather than fixed IP addresses, encryption is now the default for web traffic, and protocols such as QUIC reduce reliance on long lived client addressing by operating at higher layers.

These changes have led some to question whether IP addressing matters as much as it once did. In reality, scalable addressing still underpins everything. Data must still be routed efficiently across global networks, and infrastructure still needs predictable, manageable address allocation.

Simplifies large scale network design

IPv6 allows networks to be designed more simply at scale because large address blocks can be allocated hierarchically, which reduces routing complexity and makes networks easier to manage, particularly in data centres, content delivery networks, and emerging Internet of Things deployments where device counts can grow rapidly.

Ongoing challenges

Despite its advantages, IPv6 is not without its drawbacks. Running IPv4 and IPv6 side by side increases operational overhead because security teams must monitor and protect two protocols at once, and misconfigured IPv6 can create unexpected exposure if administrators focus only on IPv4 controls.

Some older hardware and software either lacks IPv6 support or implements it poorly, which leads organisations in those environments to disable IPv6 entirely to avoid instability, even though doing so can create long term technical debt.

IPv6 migration also requires planning, testing, and staff training, and analysts at Gartner have repeatedly noted that many organisations struggle to justify IPv6 projects without external pressure such as address exhaustion, cloud pricing models, or regulatory expectations.

Why IPv6 still matters in 2026

As the global pool of unused IPv4 addresses has become effectively exhausted, supporting new services, devices, and networks increasingly depends on complex translation layers, which are harder to scale and manage over time, while IPv6 provides a way to support continued growth without compounding that complexity indefinitely.

IPv6 was designed as underlying infrastructure rather than a visible end user technology, with its value lying in the capacity and flexibility it provides to support internet expansion as higher level protocols, encryption, and application architectures continue to evolve.

Viewed in that context, IPv6 has not failed, but has quietly fulfilled its original purpose by allowing the internet to keep growing without breaking under the strain of address scarcity and architectural workarounds.

What does this mean for your business?

IPv6’s 30 year history shows that it was never meant to be a dramatic switchover moment, but a long term safety valve for internet growth, and that role is now clearer than ever. IPv4 continues to function through layers of workarounds, trading markets, and translation systems, while IPv6 quietly carries much of the internet’s newest traffic, particularly where scale and automation matter most. The result is an internet that runs on both protocols at once, not because that was the ideal outcome, but because it proved to be the most practical one.

For UK businesses, this creates a more immediate and pragmatic challenge than a theoretical one. Organisations planning cloud migrations, rolling out new digital services, or deploying large numbers of connected devices increasingly need to understand how IPv6 fits into their infrastructure, even if customers never notice it directly. Ignoring IPv6 entirely can introduce hidden risks, from security blind spots to unexpected compatibility issues with cloud platforms and mobile networks that already assume IPv6 support by default.

For network operators, regulators, vendors, and standards bodies, IPv6 remains a reminder that core internet technologies evolve slowly and unevenly, shaped as much by economics and operational reality as by technical design. Thirty years on, IPv6 has neither transformed the internet nor been left behind by it. Instead, it seems to have become part of the underlying fabric that allows the internet to keep expanding, quietly doing the job it was built to do while the debate about its future continues.

The post How the 30 year old IPv6 still underpins the internet’s growth appeared first on Quayside Technical Services.

A new approach for detecting unintended behaviour

OpenAI’s latest research introduces what it calls a “confession”, which is a second output that sits alongside the model’s main answer. The main answer is trained in the usual way, scoring well when it is helpful, correct, safe, compliant and aligned with user expectations. However, the confession is different: it is judged only on honesty, and nothing the model says in this second output can negatively affect the reward for the first.

The idea appears to be fairly simple. When an AI system generates an answer, it is often balancing many objectives at once, which can push it towards unintended behaviour. An answer might look polished from the outside, but internally the model may have relied on a shortcut or rewarded itself for the wrong thing. The new technique tries to surface these hidden steps. As the blog puts it, the confession is meant to provide “a full account of the model’s compliance with the letter and spirit of its policies and instructions”.

The model must describe what happened

To produce a confession report, the model is asked to explain three things. It must list the instructions and constraints it believes it should have followed, assess whether its answer actually met each of them, and note any points where the instructions felt ambiguous or unclear. The aim is not to penalise the model for mistakes, but to encourage it to describe what really happened.

Why OpenAI believes confessions work

According to the research team, many types of unwanted behaviour appear because the combined reward signal used in reinforcement learning mixes too many objectives. Earlier studies found that some datasets rewarded confident but incorrect answers more than honest uncertainty. Other research showed that models could become overly agreeable when trained too heavily on preference data, or exploit loopholes in reward systems, known as reward hacking.

OpenAI argues that by treating honesty as a completely separate objective, confessions give the model a reason to admit behaviour that would normally go unnoticed. The blog states that honesty emerges because providing a truthful, evidence supported explanation is easier for the model than maintaining a consistent fabricated account that could pass scrutiny from a judging model.

Testing the method under stress

To examine the approach, OpenAI says it trained a version of GPT 5 Thinking to produce confessions, then evaluated it on a set of adversarial stress tests. The tests were specifically designed to provoke unwanted behaviour such as hallucinations, instruction violations, reward hacking, or scheming.

Across these evaluations, OpenAI reports that the probability of a false negative, where the model broke instructions and then failed to confess, averaged 4.4%. OpenAI says that, in many scenarios, the confession channel actually performed better still. When the model engaged in hacking behaviour, for example, it confessed to breaking the rules in nearly 90% of cases. In tests involving hallucinations and instruction following, most confessions correctly identified non compliance and explained why it happened.

Performance variable

OpenAI reports that performance was variable across task types, but with an average confession rate of about 74% when the model had misbehaved. In four of the 12 tests, the confession rate exceeded 90%, while in two others it fell to around 50%. OpenAI says that these differences reflect the complexity of certain tasks and the difficulty of judging compliance when instructions are ambiguous.

Example

One example published on OpenAI’s blog illustrates how the system behaves in practice. A model was asked, somewhat weirdly, to write a letter from the perspective of a chef taking up a new sport, with the added instruction to alternate sentence lengths between 12 and 18 words. The main answer attempted to follow this rule by adding indicators such as “(12)” and “(18)” at the end of each sentence, although the actual counts were not always correct. The confession then reviewed the output and acknowledged that, while it had tried to meet the requirement, it had not achieved the precise word lengths requested.

Experiments with deliberately weak reward signals

OpenAI also examined what happens when both the main reward signal and the confession reward signal are judged by a weaker model that can be easily exploited. In this setup, the main answer gradually learned how to satisfy the weak judge, producing responses that appeared to score highly while moving further away from genuine accuracy.

The confession channel behaved very differently. Despite being assessed by the same weak judge, its honesty increased as training progressed. When the model broke instructions, the confession eventually acknowledged the failure in almost every case. OpenAI sees this as evidence that, even in a flawed training environment, the most reliable way for a model to maximise its confession reward is simply to tell the truth about what it did.

How the research community is reacting

The technique has received interest from researchers who focus on AI safety and oversight. It has also prompted some scepticism. Some researchers have questioned whether any form of self reporting can be relied on in systems that already show signs of deceptive behaviour in controlled tests. They argue that a model trained to optimise for favourable outcomes may still learn ways to present itself as compliant, which raises doubts about how far confession mechanisms can be trusted on their own.

Brings bad behaviour to the surface

OpenAI acknowledges these concerns. In fact, the company stresses that confessions “do not prevent bad behaviour; they surface it”. They are meant to act as a diagnostic tool, similar in purpose to chain of thought monitoring, which attempts to reveal hidden reasoning. Both techniques aim to make internal behaviour more legible, but neither can directly stop harmful outputs on their own.

Proof of concept

OpenAI’s work on this could be described as a proof of concept. Training was conducted at a relatively small scale compared with full commercial systems, and confession accuracy remains imperfect. False positives occur, where the model confesses despite following instructions, and false negatives occur, usually because the model misunderstood the instructions or did not realise it had made a mistake.

Implications for organisations using AI

While this research is not yet part of any customer facing product, it hints at a possible direction for oversight mechanisms in future AI deployments. In theory, confession style reporting could provide an additional signal for risk teams, for example by highlighting answers where the model believes it might have violated an instruction or where it encountered uncertainty.

Industries with strong regulatory oversight may find structured self analysis useful as one component of an audit trail, provided it is combined with independent evaluation. Confessions could also help technical teams identify where models tend to cut corners during development, allowing them to refine safeguards or add human review for sensitive tasks.

Fits within a broader safety strategy

OpenAI places confessions within a broader safety strategy that includes deliberative alignment, instruction hierarchies, and improved monitoring tools. The company argues that as AI systems become more capable and more autonomous, there will be greater need for techniques that reveal hidden reasoning or expose early signs of misalignment. Confessions, even in their early form, are presented as one way to improve visibility of behaviour that would otherwise remain obscured.

What does this mean for your business?

The findings appear to suggest that confession based reporting could become a useful transparency tool rather than a guarantee of safe behaviour. The method exposes what a model believes it did, which offers a way for developers and auditors to understand errors that would otherwise remain hidden. This makes it easier to trace how an output was produced and to identify the points where training signals pulled the model in an unintended direction.

There are also some practical implications for organisations that rely on AI systems, particularly those in regulated sectors. UK businesses that must demonstrate accountability for automated decisions may benefit from structured explanations that help build an audit trail. Confessions could support internal governance processes by flagging moments where a model was uncertain or believed it had not met an instruction, which may help risk and compliance teams decide when human intervention is needed. This will matter as firms increase their use of AI in areas such as customer service, data analysis and operational support.

Developers and safety researchers are also likely to see value in the technique. Confessions provide an additional signal when testing models for unwanted behaviour and may help teams identify where shortcuts are likely to appear during training. This also offers a clearer picture of how reward hacking emerges and how different training setups influence the model’s internal incentives.

OpenAI’s framing makes it clear that confessions are not a standalone solution, and actually sit within a larger body of work aimed at improving transparency and oversight as models become more capable. The early results show that the method can surface behaviour that might otherwise go undetected, although it remains reliant on careful interpretation and still produces mistakes. The wider relevance is that it gives researchers, businesses and policymakers another mechanism for assessing whether a system is behaving as intended, which becomes increasingly important as AI tools are deployed in higher stakes environments.

The post OpenAI Trains AI Models To Confess When They Break The Rules appeared first on Quayside Technical Services.

Data spending overtakes oil for the first time

The International Energy Agency has reported that global spending on data centres will reach around $580 billion this year, overtaking the $540 billion allocated to new oil supply projects. The agency described this comparison as a clear marker of how modern economies have become anchored in digital services, cloud computing and large scale AI models, all of which require vast physical infrastructure and reliable electricity.

Usage to triple by 2035

Electricity use from data centres is projected to approximately triple by 2035. AI systems are a major driver, and the IEA expects half of all demand growth to take place in the US, with Europe and China accounting for most of the remainder. Many new facilities are located near existing clusters around large cities, with around half of the sites currently in development designed to deliver at least 200 megawatts.

The concentration of this growth is, therefore, already testing the limits of energy systems. Grid connection queues for new facilities continue to lengthen, and in several regions networks are so congested that new requests have been paused. Shortages of transformers, cables and other grid components are adding to delays. These issues highlight how the rise of AI is now tightly linked to national energy planning, rather than being a purely digital challenge.

Electricity systems under growing pressure

The IEA describes the global system as entering an “Age of Electricity”, with most new energy demand coming through power grids rather than fossil fuels. Investment in electricity generation has increased significantly since 2015, yet grid investment has not kept pace. New solar and wind capacity is being deployed at record levels, but the lines and substations needed to carry this electricity to major users are often slowed by planning processes and supply chain constraints.

Cooling demand is creating additional pressure. Rising temperatures and rising incomes in many regions are driving higher peak electricity loads from air conditioning. These peaks often coincide with the load patterns of data centres, electric vehicles and electrified heating. As a result, grids are increasingly stretched while they await new capacity and greater flexibility from storage technologies.

In several established markets, energy regulators have warned that large electricity users may need to be subject to stricter technical rules or new pricing structures to ensure network stability. Data centres are therefore becoming part of broader energy security discussions, particularly in regions where supply margins are tightening.

Power shortages slow construction across EMEA

Power constraints are directly affecting the pace of new construction across Europe, the Middle East and Africa. New research from Savills shows that only around 850 megawatts of new power capacity for data centres has been delivered across the region so far this year, representing an 11% decline compared with the same period last year. New take up has also slowed to approximately 845 megawatts, roughly half of 2024’s level.

This slowdown is not driven by falling demand. In fact, total contracted power capacity has risen to almost 14,500 megawatts, up by 12% year on year. Occupancy rates have increased to 91%, and around a quarter of new take up is now pre let. These figures illustrate that operators are securing power well ahead of time because there is no guarantee that future capacity will be available when needed.

Property advisory firm Savills found that established hubs continued to expand over the past year, including France, Germany, the UK and Ireland. Strong growth was also recorded in emerging markets such as Portugal, Saudi Arabia, Spain, the UAE and Sweden, where land and power availability are more accessible. This trend suggests that some operators are shifting attention to secondary and tertiary locations that offer fewer bottlenecks and more flexible permitting.

The effects of cost inflation

It seems that cost inflation remains a significant factor. Across EMEA, data centre build costs now range between roughly $7.3 million and $13.3 million per megawatt of IT load. It seems that some cities have even experienced double digit annual increases in land prices, labour and equipment. The result is that these rising costs are lengthening project timelines and prompting developers to form closer relationships with suppliers to secure key components earlier.

Electricity consumption forecasts continue to add urgency. One well known industry analysis last year suggested that up to 40% of data centres could face power availability constraints by 2027, and that total electricity consumption for AI optimised servers could reach around 500 terawatt hours. This would represent more than two and a half times the level recorded in 2023.

Superconductors move into data centre design

While grid upgrades are essential, many of the most immediate challenges are emerging inside existing data centre campuses. As AI systems become more computationally intensive, rack level power has risen from tens of kilowatts to around 200 kilowatts in just a few years. Some operators are now planning for 600 kilowatts per rack, and there is growing discussion of multi-megawatt rack architectures.

A US based engineering company, backed by several major technology investors including Microsoft, has now adapted high temperature superconducting cables for use within data centres. The firm’s first commercial system is designed to deliver three megawatts of low voltage power through superconducting cables cooled with liquid nitrogen to approximately -196 degrees Celsius. This cooling allows the material to carry electricity with zero loss, which in turn supports far higher power density.

The company reports that its cables require around 20 times less physical space than equivalent copper cables and can deliver power roughly five times farther within a campus. A demonstration installation has already been completed at a simulated facility, and pilot deployments at live data centres are expected next year ahead of a planned commercial launch in 2027. These technologies do not replace the need for additional grid capacity, but they allow operators to make better use of limited on site power and cooling infrastructure.

Data centres and AI companies

For data centre operators, the expansion in investment highlights both opportunity and risk. Facilities with dependable power connections, competitive energy prices and space for expansion can attract long term demand from cloud providers and AI companies. At the same time, rising construction costs, lengthy permitting and potential regulatory intervention make project planning more complex. There is increasing attention on how much electricity AI infrastructure consumes, which may influence approval processes in some regions.

It seems that AI companies now face equally important considerations. Access to high density, well powered infrastructure directly shapes the pace at which new models can be trained and deployed. Delays in securing suitable hosting capacity can slow research progress or increase operational costs. There is also growing pressure for AI to run on renewable energy, which means the location of data centres and the structure of power contracts matter more than ever.

Governments, economies and businesses

Governments now have to balance national competitiveness with energy security and climate commitments. Data centres underpin cloud services, logistics, digital payments and AI driven innovation, yet they also place significant demands on power networks. This means that policymakers must decide where new facilities can be built, how grid upgrades should be prioritised and how to maintain public support when large projects are proposed near urban areas.

Economically, the sector supports construction, engineering, manufacturing and digital roles. The long term nature of data centre contracts also encourages investment in renewable energy, battery storage and potentially small modular nuclear reactors, which several countries are exploring as a source of stable low carbon power for high demand sites.

For ordinary businesses using cloud and colocation services, the main effects are likely to be reliability, availability and cost. Capacity constraints may lead to higher hosting costs in busy regions, while areas with strong renewable resources and efficient planning may become more attractive for new deployments.

Investors and infrastructure funds increasing

Another relevant trend here is that investors and infrastructure funds continue to increase their exposure to the sector. Since 2021, around 80% to 90% of the value of closed data centre deals has involved private equity, infrastructure funds or real estate investors, compared with half in 2020. This reflects confidence in the long term demand for digital infrastructure but also raises questions about concentration of ownership in assets that underpin national digital resilience.

Challenges and criticisms

The scale of AI related electricity use has raised many questions about environmental sustainability, especially where data centres draw power from grids still reliant on fossil fuels. Concerns have been raised about water consumption for cooling, land use in crowded urban regions and the impact of construction on local communities.

Energy regulators have also highlighted system risks linked to large power users. For example, data centres can influence grid stability if they ramp up unexpectedly or disconnect suddenly, prompting discussions about new standards or pricing structures. There are wider equity concerns too, as global statistics show that hundreds of millions of people still lack basic access to electricity while trillions of dollars flow into advanced digital infrastructure.

What does this mean for your business?

The trends here show a sector that’s expanding rapidly while running up against some clear structural limits. Investment is rising because demand is strong and immediate, yet the electricity needed for large scale AI is difficult to deliver at the pace operators require. This creates a landscape where data centres are becoming essential to economic performance, but their growth is constrained by the slow evolution of energy infrastructure.

Operators now depend far more on securing reliable power than on adding floorspace or equipment. This means that sites with firm grid connections and competitive energy costs will be best placed to meet rising AI demand, while regions with slow planning processes or congested networks risk falling behind. AI companies face similar pressures because training and running advanced models depends on reliable access to powerful, energy intensive processing systems. Delays caused by grid bottlenecks or supply chain issues can slow deployment and raise operating costs.

Governments must now balance digital competitiveness with energy security and climate targets. Data centres support cloud services, logistics, payments and AI innovation, so the ability to host them is becoming a strategic priority. Grid upgrades, renewable investment and more efficient permitting processes will be required if countries want to remain competitive. This matters directly to UK businesses, which rely on stable cloud services and cost effective data processing. Rising pressure on electricity networks could influence the reliability and price of digital services across the economy.

It seems that investors are continuing to increase their involvement because long term demand remains strong, although greater private ownership of strategic infrastructure raises questions about affordability and resilience. Meanwhile, environmental concerns around electricity use, water consumption and land availability remain under close scrutiny. These issues highlight the importance of ensuring that rapid AI and cloud expansion aligns with national climate goals and local community interests.

The overall picture, therefore, appears to be that of a sector that will continue to grow but will be shaped most of all by the availability, cost and cleanliness of electricity. The choices made now on grid investment and energy policy will likely define how quickly AI infrastructure can expand and how the associated benefits are shared across economies and industries.

The post Data Centre Investment Overtakes Oil appeared first on Quayside Technical Services.

Mixed results from promising tech

The UK Department for Business and Trade (DBT) has published the results of a detailed trial of Microsoft’s M365 Copilot AI assistant, revealing no definitive evidence that the tool leads to higher productivity. Despite users reporting moderate satisfaction and perceived time savings, the trial concluded that the AI often performed inconsistently across tasks, and in some cases, reduced output quality.

The trial, which ran from October to December 2024, involved 1,000 Copilot licences distributed across the department, with roughly 300 participants consenting to monitored usage. The pilot aimed to assess the AI’s real world impact on common digital tasks using Microsoft 365 apps such as Word, Outlook, Teams, Excel and PowerPoint.

Microsoft markets Copilot as an AI productivity enhancer that can summarise meetings, draft emails, generate slides, analyse data, and more. However, the DBT report suggests the real world impact was far more nuanced than the promotional material might suggest.

Where Copilot worked and where it didn’t

The government study found that users were most satisfied when using Copilot to perform simpler, text based tasks, for example, summarising meetings, writing emails and condensing written communications.

In the trial, these tasks consistently showed time savings and improved clarity when compared with work from non Copilot users. Email writing was slightly faster and judged higher in quality and accuracy.

However, performance in more complex tasks was notably weaker. For example, data analysis in Excel and visual content creation in PowerPoint suffered, with AI-generated outputs often requiring correction or falling short of expectations. PowerPoint slide creation was seven minutes faster on average, but to a lower standard of quality. In Excel, AI users took longer and produced less accurate results than their non AI counterparts.

The report concluded: “We did not find robust evidence to suggest that time savings are leading to improved productivity. However, this was not a key aim of the evaluation, and therefore, limited data was collected to identify if time savings have led to productivity gains.”

Light usage of Copilot

The study also revealed relatively light usage patterns. According to the M365 Copilot dashboard, the average user triggered just 1.14 Copilot actions per working day across the 63 day pilot.

Word, Outlook and Teams saw the highest engagement, but more specialised tools such as Excel, PowerPoint, Loop and OneNote saw very low uptake: less than 7% of users activated Copilot in Excel or PowerPoint on any given day. Loop and OneNote usage was negligible.

These numbers raise questions about whether the value justifies the cost. UK commercial Copilot licences currently range from £4.90 to £18.10 per user per month. For large departments or enterprises, these costs could escalate rapidly, especially if many users engage with the tool only sporadically.

User attitudes and AI limitations

While 72% of participants were “satisfied or very satisfied” with Copilot, it seems that qualitative interviews suggested that much of this enthusiasm came from the novelty or perceived time saved on repetitive admin tasks. In some cases, staff used their saved time to take training courses or enjoy longer breaks, rather than focusing on higher value work.

Interestingly, a significant number of participants (22%) reported witnessing hallucinations, AI generated inaccuracies or fabrications. Another 11% were unsure, highlighting the still fragile trust in GenAI tools in professional settings.

Adoption also varied across teams, often influenced by management attitudes. Some line managers actively encouraged use, while others created a “frosty” culture around AI assistance, which in turn impacted engagement.

Microsoft and its competitors

For Microsoft, the report is clearly a mixed result. The company has heavily invested in integrating Copilot across its core product suite and has made productivity gains a central part of its pitch. But in the DBT trial, the return on investment appears questionable.

Critics say that the AI’s strengths in low complexity tasks are well documented, but the promise of broad based productivity enhancement still feels premature.

A recent MIT survey cited in the DBT report found that 95% of companies investing in generative AI tools, including M365 Copilot, had little tangible benefit to show for it. With corporate spending on GenAI already topping $40 billion, pressure is growing for vendors to demonstrate real ROI.

The findings may also embolden competitors such as Google, Zoho, or even open source productivity platforms. For now, Copilot’s core strength appears to lie in routine, text heavy admin support. In more complex tasks requiring judgement or accuracy, it remains inconsistent.

As DBT continues to analyse the environmental and cost impacts of Copilot, Microsoft may need to further refine how its AI interacts with different apps and workflows, or risk a broader slowdown in enterprise adoption.

What does this mean for your business?

The DBT trial showed that M365 Copilot could save time on routine admin, but may not provide any meaningful productivity gains across a department. For UK businesses considering using the tool or using it already, this raises some serious questions about cost effectiveness, especially where licences are purchased at scale but only lightly used. With Microsoft positioning Copilot as a flagship product, the pressure to deliver clear, measurable value will only grow.

Usage data from the trial suggests that even in a controlled, well supported environment, AI tools are far from being embedded into daily workflows. Inconsistent performance in more complex tasks, combined with ongoing concerns about hallucinations, points to a maturity gap that will be difficult to ignore. Competitors offering simpler or more focused AI products may now find space to challenge Microsoft’s all in one approach, particularly in areas where users need speed and accuracy over generalised support.

The post Government Trial Shows Few Measurable CoPilot Gains appeared first on Quayside Technical Services.

Why these two platforms are always compared

Teams and Zoom dominate the workplace communication market for good reason. They both offer video conferencing, chat, screen sharing, and integrations with third party tools, and both saw explosive growth during the remote work boom. However, beneath the surface, their purpose, strengths, and day to day feel are quite different.

Microsoft Teams is built for structured collaboration. It works seamlessly with Microsoft 365 and offers persistent chat, file sharing, and integrated tools for planning, automation, and documentation. Zoom, meanwhile, built its reputation on simplicity and video quality. It remains a favourite for meetings and webinars, especially with external audiences or users who don’t need the full Microsoft stack.

What is Microsoft Teams?

Microsoft Teams is a unified communication and collaboration platform that’s tightly integrated with Microsoft 365. It offers real time chat, voice and video calls, file sharing, and direct access to Office apps like Word, Excel and PowerPoint.

Teams is often used for:

• Daily team chat and project channels
• Scheduled and ad hoc meetings
• Collaborative editing of Office documents
• Structured access to files, wikis, and task management
• Integration with tools like SharePoint, Planner, Power BI, and third party apps

Microsoft Teams has steadily evolved since its 2017 launch and now includes Copilot AI features, webinar functionality, and even frontline workforce tools.

What is Zoom?

Zoom is a video first communication platform best known for its high quality meetings and ease of use. While it now offers team chat, phone, and whiteboard features, its core strength remains virtual meetings and webinars.

Zoom is commonly used for:

• External meetings and sales calls
• Training sessions and large webinars
• Virtual events with breakout rooms
• One to one or small group video calls
• Quick, link based meeting access with minimal setup

In recent updates, Zoom has added more advanced tools like Zoom Whiteboard, Zoom Phone, and a new AI companion for meeting summaries and smart responses.

How do they compare on key features?

Here’s how Microsoft Teams and Zoom stack up across the features that matter most to businesses:

Meetings and video quality

• Zoom consistently delivers excellent video and audio quality, even on weak connections. Its intuitive controls and meeting layouts make it a favourite for video first teams.
• Teams meetings integrate seamlessly into calendars and chats, and have improved video quality significantly. But some users still find Zoom quicker and simpler for spontaneous calls.

Chat and messaging

• Teams offers persistent, threaded conversations with clear channels, tagging, and integration with tasks and files. It’s a true digital workspace.
• Zoom’s chat is functional but less structured. It works well for simple direct messages or meeting follow ups, but lacks the project orientated features Teams excels at.

Webinars and large meetings

• Zoom leads here. It allows up to 1,000 participants in standard meetings and up to 50,000 with webinar add ons. It also includes advanced registration, polling and Q&A tools.
• Teams offers webinars too, up to 1,000 participants, but they require more setup and are less intuitive for external users.

Third party integrations

• Teams integrates deeply with Microsoft 365, as well as tools like Trello, Salesforce, Adobe Sign and hundreds more via its app store.
• Zoom also supports wide integrations including Slack, HubSpot, and Google Workspace, but it’s more focused on video features than full business workflows.

Ease of use

• Zoom’s interface is famously easy. Users can join meetings with one click and no learning curve.
• Teams takes longer to get used to. It’s powerful but some users find it cluttered or hard to navigate at first.

Feature comparison at a glance

If you’re weighing up which platform to use, this side by side feature summary gives a clear snapshot of how Microsoft Teams and Zoom stack up across the areas that matter most.

Security and Compliance

Security has become a major differentiator, and both platforms have improved significantly.

Microsoft Teams

• End to end encryption for one to one calls
• Data residency options and tenant controls
• Compliance with GDPR, HIPAA, ISO/IEC 27001, and more
• Conditional access, Microsoft Defender integration and retention policies

Zoom

• End to end encryption for meetings but it must be enabled
• Password protection, waiting rooms and participant controls
• SOC 2 and GDPR compliance
• Some advanced controls are only available on paid plans

For regulated industries, Teams tends to offer more robust compliance options out of the box, especially for firms already using Microsoft 365.

Pricing and plans

Both platforms offer free and paid versions but the value depends on what you already use.

Microsoft Teams

• Included in Microsoft 365 Business Basic (£4.90 per user per month), Standard (£10.30) and Premium tiers
• Free version available with limited features
• Enterprise plans available with advanced analytics, security, and integrations
• Great value for organisations already using Microsoft 365

Zoom

• Free plan with 40 minute meeting limit for groups
• Pro plan (£11.99 per user per month) with options for Business (£15.99) and Enterprise tiers
• Add ons for webinars, phone and extra storage
• Good standalone choice for videocentric organisations

Use cases: which tool fits which scenario?

Use Teams if:

• Your organisation is already using Microsoft 365
• You need persistent chat, document sharing and structured collaboration
• Your workflows span multiple apps and departments
• Compliance, security and data residency are high priorities

Use Zoom if:

• You need an easy, reliable video meeting platform
• You regularly host webinars, training, or external events
• You need fast setup and minimal onboarding
• You work with clients or guests outside your IT ecosystem

What are the limitations?

No platform is perfect, and while both Microsoft Teams and Zoom offer plenty of value, each has its own set of limitations that could affect how well it fits your organisation’s needs. Understanding these potential drawbacks will help you make a more informed, realistic choice.

Microsoft Teams drawbacks

• Can feel slow or cluttered, especially for new users
• Some features hidden behind higher tier licences
• Webinar and event features less polished than Zoom

Zoom drawbacks

• Limited collaboration tools beyond meetings
• Free plan has strict time and feature limits
• Requires add ons for full functionality, for example Zoom Phone, Zoom Rooms

So, which should you use?

It really depends on what your teams need most.

If you’re looking for a comprehensive, secure collaboration hub with deep Office integration, Teams is the clear winner. It’s ideal for long term project work, internal communication, and enterprise compliance. If your top priority is high quality meetings, external events, or client facing webinars, Zoom stands out for its simplicity and flexibility. It gets people into meetings fast, and it performs reliably every time. In fact, many businesses use both: Teams for day to day collaboration, Zoom for external meetings or events. That hybrid approach often delivers the best of both worlds.

The post Microsoft Teams vs Zoom: Which Is Best? appeared first on Quayside Technical Services.

Hoping to smooth the upgrade path

With support for Windows 10 set to end on 14th October 2025, Microsoft is stepping up efforts to showcase the benefits of its newer Windows 11 OS. The latest push is all about encouraging users to explore Windows 11’s built in features, from multitasking tools to biometric login, in the hope of smoothing the upgrade path and unlocking better everyday experiences.

For businesses, these seven tips may seem minor at first glance but, taken together, they may be seen to reflect a growing emphasis on usability, focus and personalisation. These are traits that can support productivity and security across both office and remote work environments.

So, what exactly are the tips, and what value do they offer?

1. Make the Start Menu your own

Microsoft’s first tip focuses on the Start Menu, which has shifted from its traditional left hand corner to a sleek, centralised layout in Windows 11. The company recommends personalising it by pinning frequently used apps, rearranging shortcuts, and resizing tiles to suit individual workflows.

For organisations standardising device setups across teams, this functionality allows for cleaner, more consistent interfaces, thereby helping users find what they need faster.

Adopting this tip could improve navigation efficiency and can be customised to match common app usage per role or department.

2. Master Snap Layouts for multitasking

Snap Layouts allow users to arrange multiple apps and windows on screen with ease. Hovering over the maximise button reveals several layout options, such as side by side or grid views.

This is particularly useful for professionals working across several platforms simultaneously, for example spreadsheets, emails, browsers and collaboration tools.

This tip could have value in reducing screen clutter and enhancing productivity for multitasking roles, even on single display setups.

3. Create a desktop for every project

Windows 11 supports multiple desktops, allowing users to group open apps and documents by task or project. For example, one desktop can house a user’s day to day admin work, while another holds presentations or creative tools.

Switching between desktops is as simple as pressing Windows + Tab, offering a smooth way to stay organised, especially in project based roles or client facing work.

This tip could help users maintain context and stay focused across varied workstreams without overwhelming a single workspace.

4. Use widgets to stay updated

Widgets provide a snapshot of useful updates like weather, news, calendar appointments and task lists, all accessible via a swipe or the Windows + W shortcut.

Microsoft pitches this as a way to streamline day to day information needs without opening multiple apps. Users can customise what they see and remove any widgets they don’t need.

Making more use of widgets could be ideal for quick, on the go updates, particularly for team members juggling schedules, appointments or news based roles.

5. Use Focus Sessions to minimise distraction

Focus Sessions, found within the Clock app, allow users to set structured time blocks for concentrated work. They sync with Microsoft To Do and reduce notifications during each timed session.

This tip supports better time management, encouraging workers to break large tasks into manageable chunks and avoid common distractions.

Using Focus Sessions could encourage better concentration habits and support mental wellbeing, particularly in hybrid or home based settings.

6. Log in with Windows Hello

Windows Hello replaces passwords with biometric authentication, for example facial recognition or fingerprint scanning, provided the device supports it.

Microsoft highlights this as a faster, more secure login method that also cuts down on password fatigue and reduces the risk of compromised credentials.

Using Windows Hello could strengthen endpoint security, simplify login workflows, and reduce password related support issues, all of which could be of real value to businesses.

7. Enable Dark Mode for comfort

Although it may sound like it, Dark Mode isn’t just a design choice, but is actually a feature designed to reduce eye strain, especially in low light settings or during extended screen use. Windows 11 offers a system wide toggle under Personalisation settings.

Users working late or on portable devices may find Dark Mode more comfortable, helping to reduce visual fatigue.

Using Dark Mode could actually improve long term comfort for screen heavy roles and can reduce energy consumption on OLED displays.

Why these tips matter

Together, these seven tips could be seen to demonstrate how Windows 11 is evolving beyond just a visual upgrade. They highlight Microsoft’s efforts to align the OS with modern working needs, prioritising comfort, personalisation and smarter workflows.

For IT managers and business leaders, they also present an opportunity to:

• Improve the onboarding experience for new Windows 11 users.
• Drive better day to day productivity through built in tools.
• Strengthen device level security with passwordless login.
• Encourage healthier digital habits across teams.

The key here is raising awareness: while these tips are all native to the operating system, few are actively promoted during setup, and many users still don’t explore them fully without guidance.

Not without challenges

Despite Microsoft’s efforts, for many people the road to better Windows 11 adoption isn’t entirely smooth.

For example, while some features, like Snap Layouts and Focus Sessions, are underused simply because users don’t know they exist, others such as Windows Hello require compatible hardware that may not be standard across all fleets. Also, while personalisation is encouraged, a lack of training or policy guidance could lead to inconsistent user experiences.

There’s also the broader challenge of change management. Businesses still on Windows 10 may face staff resistance to visual shifts like the central Start Menu or the introduction of widgets. Without clear benefits being communicated, users may default to old habits or disable new tools entirely.

For organisations managing hybrid or remote teams, this could mean greater emphasis on setup guides, onboarding materials, and hands on support during the transition.

What does this mean for your business?

Microsoft’s seven tips serve as a reminder that Windows 11 isn’t just about aesthetics, but is also about functionality, productivity, and preparing users for a more flexible, digital first way of working. The operating system is already packed with features designed to make daily tasks easier and more secure, but these benefits only actually materialise when users know where to find them and feel confident using them.

For UK businesses in particular, the timing matters. With the Windows 10 support deadline fast approaching, organisations are under pressure to modernise their desktop environments without causing unnecessary disruption. If applied well, Microsoft’s tips could ease that transition, helping teams get to grips with new interfaces and routines more quickly. In hybrid and remote contexts, where consistency and self sufficiency are key, small gains in usability and performance can translate into real operational value.

That said, the success of this guidance will depend heavily on how well it’s communicated, not just by Microsoft, but by IT teams, MSPs, and internal champions. Raising awareness is one thing but embedding new habits across diverse user groups is another. Businesses that invest in upfront user education, clear setup protocols, and device standardisation will be better placed to unlock the full value of Windows 11’s toolset.

For Microsoft, this latest push is both practical and strategic. By encouraging users to make the most of what’s already under the hood, it positions Windows 11 as a platform that’s ready for the future, not just a necessary upgrade, but a smarter way to work. Whether that message sticks will depend on how well the tech is backed by real world support, and whether users can see clear advantages in making the switch. For now, the tools are there: it’s just a question of how many take the time to use them.

The post Microsoft Highlights 7 Boosts For Windows 11 appeared first on Quayside Technical Services.

Awaiting approval

The deal, still subject to regulatory approval, signals Google’s most aggressive move yet to close the gap with rivals Amazon Web Services (AWS) and Microsoft Azure, while doubling down on multicloud and AI driven cybersecurity.

Who is Wiz and why does it matter?

Wiz, founded in 2020 by former Microsoft cloud executives, has quickly become one of the most talked about players in cloud security. The company offers an agentless, user friendly platform that helps organisations identify and remediate security risks across all major cloud providers, including AWS, Azure, Oracle Cloud, and yes, Google Cloud too.

Rather than waiting for breaches to occur, Wiz scans cloud environments continuously, maps out all assets and their connections, and flags high risk vulnerabilities in real time. Its promise is straightforward: understand your entire cloud environment, see where the threats are, and fix them before attackers can exploit them.

For example, if a developer accidentally leaves a storage bucket open to the internet or misconfigures a sensitive workload, Wiz alerts security teams, and even helps them prioritise the most critical risks. This “code to cloud” view has made it hugely popular with everyone from nimble start ups to sprawling enterprise IT teams and public sector bodies.

Results

The results speak for themselves. In just under five years, Wiz has reached an estimated $700 million annual revenue run rate, with projections it would have crossed $1 billion within a year. That makes it one of the fastest growing software companies in history and a highly attractive acquisition target.

Why is Google buying Wiz now?

Google Cloud may be strong on paper, with top tier infrastructure, AI expertise and a solid security record. However, it remains a distant third in the cloud infrastructure race, trailing AWS (30% global market share) and Microsoft Azure (21%) by a wide margin. Google Cloud sits at around 12% (Statista).

This acquisition by Google Cloud is designed to change that. By bringing Wiz into the fold, Google is looking to supercharge its credibility with enterprise customers and respond to two urgent trends in tech:

1. The surge in multicloud use. Most large organisations now use a mix of cloud providers for different workloads. That makes managing security even harder, and makes a vendor agnostic platform like Wiz essential.

2. The rising cybersecurity threat landscape. As more businesses go digital and deploy AI driven systems, the complexity of defending modern IT environments is exploding. Wiz offers a way to simplify that defence, and make it more proactive.

In a recent blog post, Thomas Kurian, CEO of Google Cloud, summed it up clearly: “Multicloud is something our customers want… Our commitment to multicloud means that new IT projects an organisation does with Google Cloud can work with their existing IT investments.”

Kurian also noted that AI is accelerating this trend: “AI architectures typically see large enterprises pool data from multiple places… This means multicloud protection is more critical than security for a centralised cache of data.”

Multicloud or bust

Interestingly, Google is not planning to make Wiz a Google Cloud only tool, and that’s no accident.

Wiz’s success has been built on its neutrality. Many of its customers don’t even use Google Cloud. If Google were to make Wiz exclusive, it could easily drive those users into the arms of competitors. That’s the most likely reason why the company has gone to great lengths to reassure customers that Wiz will continue to work seamlessly across all major cloud platforms.

In fact, as part of the deal, Google and Wiz have even earmarked an additional $1 billion retention package to keep key staff and leadership in place, including CEO Assaf Rappaport, and ensure continuity for clients.

As Wiz’s Rappaport puts it: “Wiz and Google Cloud are fully committed to continue supporting and protecting customers across all major clouds, helping keep them safe and secure wherever they operate.”

For Google, this “multicloud-first” positioning also helps defuse potential antitrust concerns. The deal comes at a time when Big Tech M&A activity is being watched closely by regulators in both the US and Europe. Emphasising openness, competition, and customer choice could make the acquisition easier to push through.

What’s in it for Google Cloud customers?

From a business standpoint, the deal could be seen as a major play to strengthen Google Cloud’s value proposition to enterprise and government buyers. By integrating Wiz into its broader portfolio, alongside Mandiant, Google Security Operations, and its AI threat intelligence tools, Google hopes to create a unified, AI optimised security platform that:

• Helps customers detect, prevent and respond to cloud threats faster.
• Lowers the cost and complexity of managing security across hybrid and multicloud environments.
• Boosts productivity of cybersecurity teams with automated tools and AI powered agents.
• Provides “measurable defence” by helping teams test and validate their own security controls.

Example

As an example of how this could work, a large bank using AWS for customer facing apps, Azure for internal services and Google Cloud for AI modelling, can now use Wiz to monitor all those environments from a single dashboard. That unified approach, backed by Google’s infrastructure and AI muscle, is what the tech giant hopes will make its cloud platform more attractive.

Wiz’s tools will also remain available via the Google Cloud Marketplace and other partner channels, giving system integrators and resellers access to a much broader toolkit for customers.

A $32 billion price tag: and plenty of pressure

At $32 billion, the acquisition is the largest in Google’s history and one of the biggest cybersecurity deals ever made. To put that into perspective, it’s more than three times the price Google paid for Motorola Mobility in 2012, and well above the $5.4 billion Amazon spent acquiring MGM Studios in 2022.

It’s also far above Wiz’s most recent private valuation, which was reportedly around $10 billion. So why the premium? Partly, it reflects the sheer scale of the opportunity. Cloud security is one of the fastest growing segments in enterprise IT. According to Gartner, global spending on cloud security is expected to top $18 billion in 2025, up from $13.5 billion in 2023.

Google sees Wiz not just as a product, but as a platform, and one that could become the default choice for securing modern cloud native and AI driven systems. The acquisition could give Google a critical edge as AI becomes more embedded in everything from healthcare to finance to national infrastructure.

However, it also raises expectations. With such a steep price tag, the pressure will now be really on Google to show real returns not just in terms of revenue, but in market share gains and customer trust.

What this means for the rest of the cloud market

For competitors like AWS and Microsoft, the acquisition of Wiz is a clear signal that Google is not backing down in the cloud wars. It’s a bet on openness, AI integration and simplified security, and it’s likely to accelerate innovation across the industry.

However, some analysts have warned that integrating Wiz successfully won’t be easy. Cultural clashes, product overlaps, and customer scepticism could all pose challenges. There are also questions around whether Wiz’s lean, start up style pace can be maintained under the umbrella of a tech giant like Google.

Still, if Google can pull it off, the payoff could be significant. As cyber threats grow more sophisticated and cloud environments become ever more complex, customers will be looking for solutions that just work, regardless of where their data lives. That’s essentially the promise Google is betting $32 billion on.

What does this mean for your business?

By snapping up one of the most agile and highly regarded multicloud security firms on the market, Google has positioned itself to offer something that many enterprise customers have long been asking for: a single, unified security platform that doesn’t demand vendor lock in and can evolve with their increasingly complex IT environments.

For Google, this is about more than plugging a gap in its offering. It’s a strategic move to become a serious contender for the next wave of cloud growth, where AI, hybrid systems and multicloud deployments are the norm. If it can successfully integrate Wiz and maintain the platform’s independence and speed of innovation, it may finally start to close the distance between itself and the cloud giants that have dominated the space for years.

For UK businesses, from financial institutions and healthcare providers, to the growing ecosystem of digital native start ups, the benefits could be substantial. Many are already juggling data across multiple platforms, wrestling with compliance demands like GDPR, and navigating increasingly sophisticated cyber threats. A more integrated, intelligent and cloud agnostic security solution could offer much needed simplification, cost efficiency and peace of mind. At the same time, UK based security consultancies, MSPs and technology partners may find new opportunities through the Google Cloud Marketplace and expanded integrations with Wiz’s tools.

However, the deal is not without its challenges. Regulatory scrutiny remains a looming question, and customers will be watching closely to see whether Google can truly preserve Wiz’s independence in practice. There’s also the risk that the sheer scale of the acquisition could dilute what made Wiz successful in the first place: its speed, focus and user friendly approach.

All that said, in a market crying out for more flexible, AI ready security solutions, this acquisition may be exactly the kind of move Google needed. Whether it pays off, and whether customers, partners and regulators will buy into Google’s multicloud pitch, remains to be seen. One thing, however, is clear: the race to secure the cloud just entered a new phase, and Google has firmly placed its bet.

The post Google’s $32 Billion Bet on Wiz appeared first on Quayside Technical Services.