HM Revenue & Customs has confirmed it is using artificial intelligence (AI) to monitor the social media accounts of suspected tax cheats, in what it says is a targeted approach aimed at tackling fraud and reducing the UK’s tax gap.

Using Algorithms

The disclosure came after recent reports in The Telegraph revealed HMRC was using algorithms to flag suspicious activity online. In response, the tax authority confirmed that AI is already playing a role in its investigative work. Officials stressed the technology is only used in criminal investigations and is not applied to the everyday taxpayer. In a statement, HMRC said AI is being deployed to help staff spot inconsistencies between declared income and publicly available social media posts, while also cutting down the amount of time spent on manual administration.

Spend Less Time on Admin

“Greater use of AI will enable our staff to spend less time on admin and more time helping taxpayers, as well as better target fraud and evasion to bring in more money for public services,” HMRC said in a statement. Officials added that the approach does not replace human decision-making, pointing to what they describe as “robust safeguards and legal oversight” around the process.

Why AI Is Being Deployed

The move reflects HMRC’s broader challenge of narrowing the UK’s tax gap, which in 2022–23 was estimated at £51 billion, or about 4.8 per cent of total theoretical tax liabilities. Of this, around £5.5 billion was attributed to tax evasion. Reducing this shortfall has become a priority for the government, which has promised that more efficient enforcement will increase revenues without raising taxes on working households.

By applying AI tools to tasks previously carried out manually, HMRC, therefore hopes to process large volumes of open-source data more quickly and accurately. Social media platforms such as Instagram, Facebook and TikTok are seen as sources of information because they often show details of people’s lifestyles that may not align with their declared income.

For example, cases have previously been reported where individuals claiming benefits on grounds of ill health were found posting publicly about participation in marathons, or where people declaring modest incomes shared photos of luxury cars and holidays. Until now, gathering such evidence relied heavily on human monitoring. AI allows patterns, anomalies and potential red flags to be detected across thousands of accounts in a fraction of the time.

Integration with Existing Systems

It’s hoped that this latest use of AI will complement HMRC’s wider analytical infrastructure, particularly its “Connect” system, introduced more than a decade ago. Connect draws together information from tax returns, banks, property records and other government databases to flag discrepancies between declared income and observed financial behaviour. Adding automated social media analysis into that mix gives investigators another channel through which to identify possible fraud.

Also To Provide Guidance To Taxpayers

It should be noted that the department has also been exploring AI for broader and less sinister purposes, such as helping taxpayers navigate the more than 100,000 pages of guidance on its website, and summarising customer service calls for advisers. However, it’s the use of social media monitoring that has generated the most public debate, not least because of its implications for privacy, accuracy and accountability.

Efficiency Gains and Cost-Cutting For HMRC

For HMRC itself, the adoption of AI in investigations is positioned as an efficiency gain rather than a cost-cutting measure. The government has already announced plans to recruit 5,500 new compliance staff, suggesting the technology will be used to augment rather than replace human expertise.

Tax specialists have noted that AI could help HMRC by pulling together information from multiple sources far more efficiently than manual checks. At the same time, they warn of potential pitfalls, including the risk of mistaken identity if the technology fails to distinguish between genuine accounts and those that are fake or hacked. Such concerns underline the importance of retaining human oversight in any decision-making process.

Focusing

From HMRC’s perspective, the argument is that automation allows investigators to focus their time where it is most valuable, i.e. complex cases that require judgment, rather than sifting through data for initial leads. The department also points to the potential to secure better returns on public spending, with AI-enabled enforcement expected to bring in billions of pounds in additional tax revenues over the coming years.

The Implications for Taxpayers

For law-abiding taxpayers, HMRC has been keen to stress that there is little to fear from AI social media monitoring. Officials say the tools are not used in routine tax collection or compliance checks, but only in serious criminal cases where there is already suspicion of wrongdoing.

The wider deployment of AI across HMRC’s operations could, in time, make interactions with the tax system simpler for the majority. For example, tools that help people understand complex rules, avoid errors in returns and access relevant guidance more quickly could reduce unintentional mistakes. Tax specialists point out that significant amounts of revenue are often lost through basic errors, and AI has the potential to reduce this by offering clearer digital assistance and more reliable support to taxpayers.

Surveillance Fears

At the same time, it has to be said that the disclosure that HMRC is scouring social media using AI may add to public concerns about surveillance and government access to personal data. Although the posts in question are publicly visible, many people may not anticipate that their online activity could form part of a criminal tax investigation.

Challenges

Not surprisingly, the announcement has not gone unchallenged. Privacy advocates and some parliamentarians have warned that reliance on AI risks producing false positives, where innocent individuals are wrongly flagged as suspicious. Concerns have also been raised that the system could amplify errors if it fails to distinguish between genuine accounts and those that are hacked, fake or misleading. Civil liberties group Big Brother Watch has previously cautioned that the spread of AI-driven surveillance represents a “frightening expansion” of state monitoring, and argued that without strict legal safeguards such tools risk undermining privacy and fairness in the justice system.

There is also a wider political backdrop. The UK government is under pressure over its broader AI strategy, with reports earlier this year suggesting the national AI institute faced internal turmoil and potential funding withdrawals. Against that context, the use of AI by HMRC has become part of a larger debate about how far public authorities should adopt emerging technologies, and how safeguards should be applied.

Some commentators have also drawn parallels with the Horizon IT scandal at the Post Office, in which faulty computer evidence led to wrongful prosecutions of sub-postmasters. Although HMRC emphasises that human decision-making remains central to its process, critics argue that overreliance on automated systems without adequate checks could carry significant risks.

The tax authority insists that its approach has been developed with those lessons in mind. AI is used to highlight possible leads, but human investigators must review the evidence before any action is taken. Officials have also stressed that all use of AI in this area is subject to legal oversight, ensuring compliance with UK data protection and criminal justice laws.

Wider International Context

It should be noted, however, that the UK is not alone in deploying AI to tackle tax evasion as other countries have already been using similar methods. HMRC has been sending so-called “nudge letters” to taxpayers after analysing international financial data shared under the OECD’s Common Reporting Standard, which tracks overseas income. This AI-assisted approach has contributed to a reported 22 per cent rise in admissions of foreign tax evasion in recent years.

The logic appears to be that as more financial data becomes digitised and globally shared, AI will become a necessary tool for connecting the dots. For HMRC, adding social media monitoring into that picture is seen as the next step in keeping pace with increasingly complex forms of tax fraud.

What Does This Mean For Your Business?

The real test will be whether HMRC can strike the balance between efficiency and fairness. The technology may promise to sharpen investigations, close the tax gap and support a more resilient tax base, but it also raises unresolved questions about how far government agencies should be able to probe into people’s digital lives. For taxpayers, reassurance that AI will not be used in routine compliance checks will be welcome, but concerns about transparency and proportionality are unlikely to fade quickly.

For UK businesses, the implications may be twofold. For example, firms could benefit from a more level playing field if AI enables HMRC to clamp down more effectively on undeclared income and fraudulent competition. However, there is a risk that legitimate companies could face greater scrutiny or reputational harm if mistakes occur in the interpretation of online data. The safeguards HMRC has promised will, therefore, be critical in maintaining trust.

Other stakeholders, from policymakers to civil liberties groups, will, no doubt, be watching closely. The government has positioned AI as a tool for improving public services and raising revenue without additional tax rises, yet its wider strategy for regulating AI remains unsettled. With memories of past technology failures still fresh, the rollout of these systems within HMRC will be an early indicator of how the state intends to manage the trade-offs between innovation, privacy and accountability in the years ahead.

The UK Government has announced the deployment of 10 new Live Facial Recognition (LFR) vans across seven police forces in England, saying the move will help officers track down suspects wanted for some of the most serious crimes.

What?

On 13 August, the Home Office confirmed that vans equipped with live facial recognition cameras will be shared between Greater Manchester, West Yorkshire, Bedfordshire, Surrey and Sussex (jointly), and Thames Valley and Hampshire (jointly). The rollout is part of the government’s wider “Plan for Change” in policing, which also includes neighbourhood policing guarantees and the recruitment of 13,000 more officers by 2029.

The new vans are expected to be deployed in the coming weeks. They will be coordinated nationally by the National Police Chiefs’ Council and South Wales Police, which has already led the way in using the technology. The government says each vehicle will be staffed with trained officers who verify every potential match generated by the software.

How Does the Technology Work?

Live facial recognition compares the faces of people passing by with those held on a police “watchlist”. These watchlists are drawn up for each deployment and include individuals wanted for offences such as sexual assault, violent crime, murder, and breaches of bail or court conditions.

The technology measures facial features such as the distance between a person’s eyes or the shape of the jawline, creating a digital signature that can then be matched against the watchlist. If a possible match is made, an officer on site reviews the image before deciding whether to stop the individual.

Why Is the Government Expanding It Now?

Home Secretary Yvette Cooper said the expansion reflects both the need to modernise policing and the results already seen in London and South Wales. “Facial recognition will be used in a targeted way to identify sex offenders or people wanted for the most serious crimes who the police have not been able to find,” she explained, adding that the rollout will be backed by a new legal framework and public consultation on safeguards.

Only Intelligence-Led Use

It’s worth noting here that the government says that officers will only use the technology in intelligence-led deployments, and that strict guidance from the College of Policing and the Surveillance Camera Code of Practice will apply. It insists the aim is not mass surveillance but a proportionate use of new tools to protect the public.

What Have Previous Trials Achieved?

The Metropolitan Police and South Wales Police have been the two leading forces trialling LFR technology. According to the Met, 580 arrests were made in London over a 12-month period up to mid-2025, covering offences including rape, domestic abuse, robbery, and knife crime. Among those detained were 52 registered sex offenders who had breached their conditions.

South Wales Police has emphasised its record, reporting no false alerts since August 2019. Chief Superintendent Tim Morgan has argued the technology has been deployed ethically and lawfully, saying it has “never resulted in a wrongful arrest” in the force’s area. It has been used at large public gatherings such as football matches and concerts, where suspects for serious offences have been identified and detained.

However, these claims have not gone unchallenged. For example, earlier deployments in South Wales were the subject of a legal challenge brought by a Cardiff resident (Ed Bridges), who argued the technology was used unlawfully to scan him in public without consent. Although the Court of Appeal ruled in 2020 that South Wales Police’s use of LFR was unlawful on three grounds, including a lack of proper safeguards against arbitrary use, the force was later cleared to continue under revised guidance.

The Metropolitan Police has also faced criticism after individuals (including Londoner Shaun Thompson) were wrongly flagged by cameras during live deployments. Civil liberties groups say such errors highlight the risks of false positives, even if no wrongful arrests ultimately followed.

Examples of use include London deployments in violent crime hotspots and at high-profile events such as a Beyoncé concert, where police used the technology to search for individuals suspected of terrorism or child abuse offences. While supporters say these trials show the technology’s potential to locate dangerous offenders in crowded settings, opponents argue the risks to privacy and the possibility of misidentification remain significant.

The Case for Benefits

Proponents argue that live facial recognition allows officers to locate suspects more quickly than traditional policing methods. For example, Lindsey Chiswick, the National Police Chiefs’ Council lead for facial recognition, said the new vans provide “an excellent opportunity for policing”, with every deployment being “targeted, intelligence-led, within a set geographical location and for a defined period of time”.

The Tony Blair Institute for Global Change has also supported its use, describing it as “a no-brainer” for catching violent offenders in crowded public places. They argue that those not on a watchlist will have their faces pixelated, with no data retained.

For police forces, the vans are expected to essentially act as a force multiplier, freeing up resources and providing a visible deterrent at a time when officers are under pressure to deal with rising violent crime and anti-social behaviour.

Criticism and Concerns

Civil liberties groups remain strongly opposed to the use of this technology. For example, Big Brother Watch described the rollout as “a significant expansion of the surveillance state” and has lodged a legal challenge against the Metropolitan Police’s use of the technology. Interim director Rebecca Vincent argued that police were treating the lack of a specific legal framework as “carte blanche” to expand the practice without democratic scrutiny.

Also, Baroness Chakrabarti, former director of Liberty, called the technology “incredibly intrusive” and warned that it posed risks to privacy, freedom of assembly, and the possibility of false matches. She welcomed the consultation but criticised the absence of legislation to date, saying the technology had so far been deployed “completely outside the law”.

Amnesty International UK has also criticised the move, describing LFR as “dangerous and discriminatory”. The group warns that even if algorithms are shown to be unbiased in laboratory tests, real-world deployments risk disproportionately targeting minority communities.

The Information Commissioner’s Office (ICO), which regulates data protection, emphasised that facial recognition “does not operate in a legal vacuum” and must always be lawful, fair, and proportionate. The ICO is due to publish its findings on the use of the technology by South Wales and Gwent Police.

What Does Independent Testing Show?

To address bias concerns, the government commissioned the National Physical Laboratory to test the algorithms being used in the vans. The laboratory reported that at the settings used by police, the system was accurate and showed no bias by age, gender, or ethnicity.

Even so, critics argue that independent testing cannot address all the risks. The possibility of false positives, mission creep into less serious crimes, or erosion of trust in policing remain live debates.

Implications for Communities and Businesses

For the public, the presence of LFR vans is likely to be visible in town centres and at large public events. The Home Office has said that signage will make it clear when the technology is being used, and that data will only be retained for the period of the deployment.

Businesses in affected areas may see benefits if the vans deter shoplifting, robbery, or violent crime in retail and nightlife districts. However, there are concerns about potential impacts on footfall if people feel uncomfortable being scanned while shopping or socialising.

Neighbourhood organisations such as Neighbourhood Watch have broadly welcomed the balance between more officers and targeted technology. Chief Executive John Hayward-Cripps said the combination of “responsive policing that prioritises local relationships” with modern tools could help rebuild public trust.

For police officers themselves, the vans offer a new way to match intelligence with on-the-ground enforcement, but they also come with added responsibilities for transparency, proportionality, and accuracy.

What Does This Mean For Your Business?

The government’s decision to expand live facial recognition vans puts advanced surveillance tools directly into mainstream policing, but it also raises difficult questions about oversight. For police forces, the benefits are obvious, i.e., the vans make it easier to identify wanted suspects in crowded areas and allow officers to focus on other frontline duties. The record of arrests in London and South Wales shows the technology can be effective against serious offenders, yet the court rulings and cases of mistaken identification show why critics say it cannot be deployed without stronger safeguards.

For the public, the experience is likely to be mixed. For example, some will welcome the reassurance of extra tools against violent crime, while others will be uneasy at the idea of their faces being scanned in public. Public trust will hinge on whether officers use the vans in line with clear rules and whether those rules are seen to be independently enforced. Businesses in city centres may gain if the technology reduces theft, disorder and violence near shops and venues, although any perception of over-policing could also deter customers.

The government’s consultation on new legislation is, therefore, critical. Campaigners argue that police have been using LFR outside a proper legal framework for years, while ministers insist deployments will stay targeted and proportionate. The outcome will decide whether this expansion strengthens policing while protecting rights, or whether it fuels a deeper public backlash that undermines confidence across communities, businesses and other stakeholders.

Elon Musk’s Tesla has applied for a licence to supply electricity to British homes and businesses, a move that could see the US-based firm directly enter the UK’s highly regulated energy market from 2026.

What Has Tesla Done?

The application was formally submitted on 25 July by Tesla Energy Ventures Limited, a UK-registered company under the wider Tesla umbrella. Ofgem, the UK’s energy regulator, confirmed the licence request on its website and set a consultation period until 22 August for stakeholders to submit comments. In line with the Utilities Act 2000, Ofgem has only published the notice of application, limiting further detail. If approved, the licence would allow Tesla to operate as a retail electricity supplier across England, Scotland and Wales.

Extend Scope of Existing Licence

It should be noted here that Tesla already holds a generation licence in the UK (granted in 2020) which allows it to produce electricity. However, this latest move could extend its scope to selling power directly to households and businesses, in the same way as established suppliers such as British Gas, Octopus Energy and OVO.

Why Is Tesla Doing This?

Tesla’s decision comes at a time when the company is facing declining electric vehicle sales across Europe. UK registrations in July fell by almost 60 per cent compared to the previous year, while German sales dropped by 55 per cent. Across ten key European markets, Tesla’s sales were down by 45 per cent. Competition from rival EV manufacturers, particularly China’s BYD, has been a major factor in this slump.

Diversifying into energy supply, therefore, offers Tesla another route to growth, particularly given its sizeable existing footprint. More than 250,000 Tesla vehicles are on UK roads and tens of thousands of its Powerwall home battery systems have been installed. This customer base could provide a ready pool of households willing to adopt Tesla’s electricity supply, particularly if bundled with discounts for charging vehicles or exporting stored solar energy back to the grid.

Like In Texas

The company has already built up experience as an energy supplier in the United States. In Texas, Tesla Electric launched in 2022 as a retail provider offering households low-cost, 100 per cent renewable power. Customers do not need to own a Tesla product to join, though EV owners and Powerwall users are offered cheaper charging rates and the ability to sell surplus electricity back to the grid.

The Texas operation also supports the concept of a “virtual power plant”, where thousands of home batteries are linked together to provide grid stability. Tesla has suggested that similar models could eventually be deployed in other markets. In the UK, this would align with National Grid’s push for more flexible energy resources and time-of-use tariffs that encourage households to use power at off-peak times.

What Could This Mean for the UK Market?

If the licence is granted, Tesla would join a market that is both competitive and tightly controlled. The so-called “Big Six” suppliers, now expanded to include Octopus alongside British Gas, EDF, E.ON, OVO, ScottishPower and SSE, still dominate with more than 90 per cent of the domestic supply market. Smaller and newer entrants have struggled in recent years, especially during the energy crisis, which saw dozens of challenger firms collapse under pressure from soaring wholesale prices.

Some analysts have pointed out that Tesla is entering a highly regulated market where profit margins are already thin and most of the big suppliers have invested heavily in smart tariffs, making it difficult for new players to break through. However, others have highlighted how Tesla’s existing ecosystem could help it stand apart. For example, despite falling EV sales, Tesla still has a sizeable footprint in the UK, with more than 250,000 cars sold and thousands of Powerwall batteries installed. That existing customer base could give Tesla a natural advantage if it follows the same model as its Texas business, where households are offered cheaper charging and paid for feeding power back into the grid.

Potential Tariff Innovation

One area where Tesla may compete effectively is in smart tariffs for EV charging and home energy storage. Between 2020 and 2023, Tesla partnered with Octopus Energy on the Tesla Energy Plan, a smart import-export tariff that allowed customers with solar panels and Powerwalls to buy and sell electricity at the same rate. Although Tesla later withdrew from the partnership, Octopus continues to offer a similar tariff, demonstrating demand for such arrangements.

If Tesla can combine its EVs, batteries and potential supply licence into a single integrated offer, it could appeal strongly to existing customers. For example, discounted tariffs for charging Teslas overnight, coupled with payments for sending energy back to the grid from a Powerwall, would create a closed-loop system that few other suppliers could match.

Challenges

Despite the potential, the barriers are considerable. The UK retail electricity market is crowded, margins are slim, and switching rates are low compared with the period before the energy crisis. Many households are locked into dual-fuel contracts that combine gas and electricity, which may make a Tesla-only electricity offer less attractive.

There is also the question of public perception. Elon Musk’s increasingly political public profile has drawn strong criticism in Europe and the UK. Also, he has described Britain as a “police state” and criticised asylum and migration policies. His closeness to US President Donald Trump (although they have since fallen out) and his actions with DOGE in the US have further polarised opinion and appear to have caused huge damage to his personal brand (and his vehicle sales). In fact, some consumer groups have warned that Musk’s views could influence whether households are willing to sign up for Tesla-branded energy.

Tesla itself has remained quiet on its application. The notice submitted to Ofgem was signed by Andrew Payne, Tesla’s head of energy for Europe, the Middle East and Africa. No public statement has been issued by the company, which has said only that it continues to expand its energy services globally.

What Does This Mean For Your Business?

Tesla’s bid to supply electricity in the UK sets up a clear test of whether its brand strength and integrated technology can overcome the realities of a tightly controlled market. On paper, its combination of cars, home batteries and solar solutions could give it an edge in offering customers genuinely joined-up energy services. In practice, it faces the same pressures that have squeezed margins for existing suppliers, alongside the added complication of public sentiment about its founder.

For UK households, the offer of cheaper EV charging or the ability to trade surplus solar power back to the grid would be attractive, particularly at a time when energy bills remain under close scrutiny. For businesses, Tesla’s entry could bring new tariff models for fleets or for sites already investing in renewable generation and storage. If the model mirrors what has been developed in Texas, it may also open the door for companies to participate in virtual power plants that improve resilience and provide income streams from energy flexibility.

For the energy sector, the move signals that disruption could just as easily come from a technology giant as from a nimble start-up. Incumbent suppliers will be watching closely, both for the pricing strategy Tesla adopts and for the way it leverages its hardware base to win loyalty. Regulators, meanwhile, will have to balance innovation with consumer protection in a market that has already seen waves of supplier failures.

Tesla is attempting to diversify at a time when its automotive business is under pressure, and the UK market will be an early test of whether energy supply can deliver the growth it now seeks. If it succeeds, it could accelerate the shift towards more dynamic and decentralised energy systems. If it fails, it will underline how difficult it remains to challenge the dominance of established players in one of Europe’s most heavily regulated markets.

AI start-up Perplexity has made a surprise $34.5 billion bid for Google’s Chrome browser, a move that has shocked the tech industry and raised questions about antitrust pressure, corporate ambition, and whether such a deal could ever succeed.

Perplexity

Perplexity is a San Francisco-based company founded in 2022 by Aravind Srinivas, a former Google and OpenAI researcher. Despite being just three years old, the firm has already become one of the highest-profile challengers in the generative AI space. Its core product is a real-time AI-powered search engine that provides conversational answers with source links, setting it apart from rivals like OpenAI’s ChatGPT and Google’s Gemini.

The company has attracted major backing, including funding from Amazon founder Jeff Bezos, chipmaker Nvidia, and Japan’s SoftBank, and was valued at around $18 billion as recently as July. Perplexity says it now serves around 30 million monthly active users and has partnerships with publishers such as Time and the Los Angeles Times.

Last month, it launched Comet, an AI-native browser designed to help users summarise web content, manage tabs more intelligently, and automate tasks such as scheduling and online shopping. Comet combines local device processing with cloud-based models, including GPT-5, Anthropic’s Claude, and Google’s own Gemini. Srinivas has described it as a “cognitive operating system” rather than a conventional browser.

The Offer For Chrome

On 12 August, Perplexity confirmed that it had made a formal $34.5 billion all-cash offer to acquire Chrome, the world’s most widely used browser with an estimated three billion users. The offer was unsolicited, and Alphabet, Google’s parent company, has not indicated any willingness to sell Chrome.

Perplexity framed the bid as an effort to preserve the principles of the open web, pledging to keep Chromium, the open-source foundation of Chrome that also underpins Microsoft Edge and Opera, fully maintained. It also promised to invest $3 billion over two years in development and to retain Google as the default search engine inside Chrome, while allowing users to change provider more easily.

Why Make This Move Now?

The timing is no accident. Google is awaiting the outcome of a major US antitrust case that found it had unlawfully monopolised the search market by locking in default search agreements with device manufacturers and browser developers. Judge Amit Mehta is due to decide on remedies, and one of the Justice Department’s proposed options is that Google could be forced to divest Chrome.

By making its bid public now, therefore, Perplexity is positioning itself as a ready buyer if regulators force a sale. The company has argued that moving Chrome into independent ownership would be in the “highest public interest” and would protect user choice.

It also reflects a broader shift in how browsers are seen in the AI era. For example, with the rise of conversational search, browsers have regained importance as gateways to traffic, data, and default search settings. Controlling Chrome, which has more than 60 per cent of the global browser market, could give Perplexity a huge distribution advantage for its own AI search technology.

Could It Succeed?

It appears that many analysts remain highly sceptical about Perplexity’s chances of success. For example, Perplexity’s own valuation is only about half the size of the bid it has put forward, and while the company has claimed that multiple funds are willing to finance the deal, it has not named them.

Estimates of Chrome’s true value vary, with some industry figures suggesting it could be worth $50 billion or more, far higher than Perplexity’s offer. Even if regulators order a sale, Google has said it would appeal, warning that divestiture would damage innovation, user privacy, and security. Legal experts believe that a final decision could take years to resolve, possibly extending through appeals to the Supreme Court.

For now, Google has given no indication it would entertain the bid. Alphabet executives have consistently argued that Chrome is central to its strategy, not just as a browser but as a linchpin connecting billions of users to its search engine and advertising business.

What Would It Mean for Perplexity?

If Perplexity somehow succeeded, acquiring Chrome would instantly transform it from a rising AI challenger into a dominant player in the internet ecosystem. Chrome’s reach would provide an unrivalled platform to push its AI search engine, while the promise to maintain Google as the default search provider could ease antitrust scrutiny.

It would also allow Perplexity to leapfrog competitors like OpenAI, which has been working on its own AI browser, and other search challengers such as DuckDuckGo. Integrating Chrome’s scale with Perplexity’s AI could accelerate the shift toward AI-native browsing, potentially changing how users interact with information online.

However, absorbing Chrome’s three billion users would also pose huge operational and financial challenges. For example, maintaining a browser of that size requires vast infrastructure, security resources, and compliance mechanisms. For a start-up, even one backed by high-profile investors, this would be an extraordinary undertaking.

What Would It Mean for Google?

For Google, losing Chrome would, of course, be a major blow. The browser is not only a product in its own right but also a strategic entry point for its search business, helping to sustain its advertising revenue. Chrome also serves as a platform for integrating new AI features such as generative overviews within search results.

Without Chrome, Google would be forced to rely more heavily on agreements with rival browsers and devices, weakening its distribution power. That is precisely why regulators see divestiture as one of the strongest remedies to restore competition in the search market.

Reactions and Criticisms

Reactions to the bid have been mixed. For example, some industry observers view it as a bold statement of intent from a company looking to define the future of web search. Others see it as unrealistic, pointing out that Perplexity has not disclosed firm financing and may simply be using the offer to raise its profile.

Scepticism has also focused on whether Perplexity’s promise to keep Google as the default search engine would truly benefit competition, or whether it is more about reassuring regulators. Critics argue that Chrome’s value lies not only in its codebase but also in its role as a pipeline for search and advertising data, which cannot easily be separated from Google’s wider ecosystem.

Business users are also likely to be watching closely. Chrome is the default browser in many workplaces, chosen for its compatibility, speed, and integration with enterprise tools. Any change of ownership could raise questions about data handling, support, and continuity, particularly if Perplexity sought to integrate more AI-driven features.

For now though, the move has generated significant debate about the future of browsers, competition, and AI in the digital economy, regardless of whether the deal itself has any chance of success.

What Does This Mean For Your Business?

Perplexity’s bid may be as much about timing and visibility as it is about ownership. With Google under pressure in the US courts, the possibility of a forced sale has moved from theory to something regulators may genuinely consider. By stepping forward now, Perplexity is signalling that it wants to be part of the solution if Chrome is prised away from Google, even if the financial and legal obstacles make that outcome unlikely in the near term.

For UK businesses, the implications are not trivial. For example, Chrome dominates the workplace browser market, and any change in ownership could alter how security, updates, and AI integrations are delivered. A Perplexity-controlled Chrome might accelerate the introduction of AI-native features into day-to-day browsing, offering new efficiencies but also raising fresh questions about data handling and reliance on emerging players. IT teams and decision-makers would need to weigh the potential benefits of AI-powered functionality against the stability that comes with Google’s long-established stewardship.

Investors and regulators will also be watching carefully. If a start-up with a fraction of Google’s size and revenue can credibly put forward an offer of this scale, it signals how central browsers have once again become in the race to dominate AI-driven search. It also shows how antitrust scrutiny is changing the competitive landscape, opening the door to bids and proposals that would once have been unthinkable.

In practical terms, Google’s strong resistance and the length of any appeals process mean Chrome is unlikely to change hands quickly. However, the bid alone has reframed the conversation about the browser’s future. Whether as a serious takeover attempt or as a calculated move to raise its profile, Perplexity has forced the industry to imagine what a post-Google Chrome might look like, and to consider how that could reshape competition, innovation, and user choice across the digital economy.

Wikipedia has lost its High Court challenge against the UK’s Online Safety Act, a ruling that could have far-reaching implications for how the online encyclopaedia operates in Britain and how the wider internet is regulated.

What Was the Challenge About?

The case was brought by the Wikimedia Foundation, the non-profit organisation that runs Wikipedia, alongside a UK-based volunteer editor. Their legal challenge focused on a specific part of the Online Safety Act known as the Categorisation Regulations. These rules decide which websites and online services fall under the law’s most stringent requirements, known as Category 1.

Category 1 platforms face additional obligations designed to curb harmful online content. For example, they must put in place systems to prevent children from accessing illegal or harmful material, enforce stricter moderation policies, and verify the identities of contributors. For Wikipedia, which relies on anonymous and pseudonymous volunteers to write and update its 65 million articles across 300 languages, this requirement was seen as fundamentally incompatible with how the site functions.

Wikimedia argued that being classified as Category 1 would undermine the privacy of its 260,000 active volunteer editors worldwide, exposing them to risks ranging from harassment to potential prosecution in authoritarian countries. The Foundation also warned that complying with such rules would require major changes to Wikipedia’s open model and could restrict access for UK users, either by cutting off large parts of the site or disabling core editing functions.

Wikipedia’s Challenge Dismissed

The case was heard at the High Court of Justice in London in July, with the ruling delivered on 11 August. Mr Justice Johnson dismissed the challenge, ruling that the Secretary of State for Science, Innovation and Technology had acted lawfully when deciding how the regulations should be applied.

The judge rejected Wikimedia’s claim that the rules were inherently irrational or too broad. However, the judgment stopped short of granting the government and Ofcom (the UK’s communications regulator) a “green light” to impose obligations that would significantly damage Wikipedia’s operations. Instead, the court emphasised that regulators still have a duty to ensure that public interest platforms such as Wikipedia are not unfairly harmed as the Act is implemented.

Wait Until Next Year

Importantly, Ofcom has not yet confirmed whether Wikipedia will be designated as a Category 1 service. That decision is expected later this year. The court noted that if Wikipedia is classified in this way and can no longer operate effectively, Wikimedia may have grounds to bring a fresh legal challenge.

What Happens Next?

For now, Wikipedia remains accessible to UK users as normal. However, the outcome of Ofcom’s categorisation process will be critical. If the encyclopaedia is placed in Category 1, the Foundation would face the unenviable choice of either overhauling its model by introducing identity checks for editors and stricter content controls or risking fines and potentially being blocked in the UK.

Will Keep Engaging With Ofcom and the Government

Wikimedia has said it will continue engaging with Ofcom and the government to find a workable solution. The Foundation also stressed that it views the ruling as a signal that regulators must tread carefully in applying the Act. The judgment recognised the “significant value” of Wikipedia and the risks to human rights if its contributors were forced to give up anonymity.

The UK government, however, welcomed the decision, arguing that the Act is essential for creating a safer online environment. Ministers have repeatedly said the law is not intended to target smaller or non-profit platforms, but rather to hold the largest and most influential services accountable.

Why Does This Matter for Wikipedia and the Internet?

At its core, this case reflects the tension between child safety and free knowledge online. Wikipedia is the fifth most visited website in the world, drawing more than 15 billion views every month. In the UK alone, its content is accessed hundreds of millions of times per month, including through school curricula such as the Welsh-language version, which is the most popular Welsh website globally.

The concern is that sweeping rules designed with large commercial platforms in mind, such as social media networks or adult content sites, could inadvertently capture collaborative knowledge projects. By forcing identity verification, the law could undermine the very openness and volunteer-driven nature that has made Wikipedia one of the most trusted sources of information.

Digital rights organisations have warned that the Act’s broad scope risks collateral damage. Groups such as the Electronic Frontier Foundation and the Open Rights Group have argued that blanket obligations may erode free speech, privacy, and editorial independence. They note that anonymity is not just a shield for trolls, but a lifeline for vulnerable people, including dissidents, journalists, and LGBTQ+ individuals who may otherwise be unable to contribute safely.

What Do Privacy and Child Safety Advocates Say?

Privacy campaigners have said that forcing contributors to hand over identity documents or undergo verification could expose them to data breaches or surveillance. In the worst cases, it could make volunteers targets for legal action in countries with harsh censorship laws. For example, Robin Wilton of the Internet Society has warned that weakening online anonymity often “exposes everyone, not just bad actors”.

Child Safety Charities Pleased

On the other side of the debate, child safety charities have welcomed the Act, arguing that the internet has long lacked accountability. They believe stronger rules are necessary to protect young people from harmful material, ranging from self-harm forums to violent pornography. These groups argue that platforms with significant reach and influence (whether commercial or not) must share responsibility for ensuring that harmful content is not easily accessible.

The Wider Implications

The High Court’s decision has sparked broader debate about the future of the internet in the UK and beyond. For example, if Wikipedia is required to comply with obligations designed for commercial giants, it raises questions about whether other non-profit or community-led platforms could face similar challenges. Some small forum operators have already shut down message boards, citing an inability to pay for age verification systems or the risk of large fines.

For businesses, the ruling highlights the increasing complexity of operating online services in the UK. Companies that rely on Wikipedia as a knowledge resource, including schools, libraries, and media outlets, could face disruption if the site is restricted. More broadly, the case demonstrates how regulatory approaches intended to improve online safety may reshape the digital environment for all stakeholders, potentially reducing openness and participation in favour of stricter control.

In practical terms, Ofcom now faces the task of balancing the government’s safety objectives with the need to protect platforms that serve the public interest. The regulator’s decisions later this year will determine not only Wikipedia’s future in the UK but also how flexible the Online Safety Act can be in practice.

What Does This Mean For Your Business?

What happens next will depend heavily on how Ofcom interprets its role. The court appears to have placed the responsibility squarely on the regulator to ensure that the Act is not applied in ways that damage public-interest resources such as Wikipedia. If Ofcom takes a strict line and categorises the site as Category 1, it would trigger requirements that may be unworkable for a volunteer-driven encyclopaedia. If it takes a more flexible approach, or recommends changes to Parliament, it could preserve both the goals of online safety and the principle of open access to knowledge.

For UK businesses, the outcome is likely to be more than a theoretical concern. For example, many companies, particularly in research, education, media and publishing, depend on Wikipedia as a readily accessible knowledge base. Any disruption to the platform’s availability could make training, communications and fact-checking more difficult. Schools and universities would also be affected if the site was restricted or stripped of key features. For firms operating online platforms themselves, the case is a reminder that compliance burdens are growing and that regulation designed for large tech firms can have unintended knock-on effects across the wider economy.

Campaigners argue that the stakes go further still. If collaborative projects such as Wikipedia are forced into compliance regimes built for social media giants, smaller online communities will almost certainly struggle to survive. That risks narrowing the diversity of online voices and discouraging innovation. On the other hand, advocates for stronger safety measures insist that the internet cannot continue to operate on principles designed in an earlier era, and that accountability must be applied across the board if young people are to be protected from harmful content.

The balance now lies in how regulators and government choose to enforce the Act. The High Court has made clear that there are limits to what can reasonably be demanded of platforms like Wikipedia. Whether those limits are respected will shape not only the future of the world’s largest encyclopaedia, but also the broader character of the internet in the UK.

Average ransomware payments have more than doubled in the past quarter, while stolen credentials are driving a sharp rise in breaches.

Coveware by Veeam reported the average ransom payout in Q2 2025 hit $1.13 million, up 104 per cent on the previous quarter. The median payment also doubled to $400,000, with larger organisations paying out in data exfiltration-only incidents, where files are stolen rather than systems encrypted. Data theft was a factor in 74 per cent of cases.

Criminal groups such as ‘Scattered Spider’, ‘Silent Ransom’ and ‘Shiny Hunters’ are using targeted social engineering to impersonate staff, trick helpdesks, and exploit third-party providers. “Attackers aren’t just after your backups – they’re after your people, your processes, and your data’s reputation,” warned Coveware CEO Bill Siegel.

At the same time, Check Point found credential theft has surged 160 per cent in 2025, now causing one in five breaches. Many businesses take months to revoke exposed logins, giving attackers time to exploit them.

Security experts advise organisations to enforce multi-factor authentication, tighten password policies, and train staff to spot social engineering. Treating stolen credentials and data theft as primary risks is now seen as essential.