UK regulators and the Treasury face MP criticism over their approach to AI, amid warnings of risks to consumers and financial stability. A new Treasury Select Committee report says authorities have been overly cautious as AI use rapidly expands across financial services.
More than 75% of UK financial firms are already using AI, according to evidence reviewed by the committee, with insurers and international banks leading uptake.
Applications range from automating back-office tasks to core functions such as credit assessments and insurance claims, increasing AI’s systemic importance within the sector.
MPs acknowledge AI’s benefits but warn that readiness for large-scale failures remains insufficient. The committee urges the Bank of England and the FCA to introduce AI-specific stress tests to gauge resilience to AI-driven market shocks.
Further recommendations include more explicit regulatory guidance on AI accountability and faster use of the Critical Third Parties Regime. No AI or cloud providers have been designated as critical, prompting calls for stronger oversight to limit operational and systemic risk.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Apple has issued a renewed warning to iPhone users, urging them to install the latest version of iOS to avoid exposure to emerging spyware threats targeting older versions.
Devices running iOS 26 are no longer fully protected by remaining on version 18, even after updating to the latest patch. Apple has indicated that recent attacks exploit vulnerabilities that only the newest operating system can address.
Security agencies in France and the United States recommend regularly powering down smartphones to disrupt certain forms of non-persistent spyware that operate in memory.
A complete shutdown using physical buttons, rather than on-screen controls, is advised as part of a basic security routine, particularly for users who delay major software upgrades.
While restarting alone cannot replace software updates, experts stress that keeping iOS up to date remains the most effective defence against zero-click exploits delivered through everyday apps such as iMessage.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
OpenAI’s annualised revenue has surpassed $20 billion in 2025, up from $6 billion a year earlier. The company’s computing capacity and user numbers have also continued to grow.
The company recently confirmed it will begin showing advertisements in ChatGPT to some users in the United States. The move is part of a broader effort to generate additional revenue to cover the high costs of developing and running advanced AI systems.
OpenAI’s platform now spans text, images, voice, code, and application programming interfaces. CFO Sarah Friar said the next phase of development will focus on agents and workflow automation that can operate continuously, retain context over time, and take action across multiple tools.
Looking ahead to 2026, the company plans to prioritise what it calls ‘practical adoption’, with a particular emphasis on health, science, and enterprise use cases. The aim is to move beyond experimentation and embed AI more deeply into real-world applications.
Friar also said OpenAI intends to maintain a ‘light’ balance sheet by partnering with external providers rather than owning infrastructure outright. Contracts will remain flexible across hardware types and suppliers as the company continues to scale its operations.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
A new ‘Answer Now’ button has been added to Gemini, allowing users to skip extended reasoning and receive instant replies. The feature appears alongside the spinning status indicator in Gemini 3 Pro and Thinking/Flash, but is not available in the Fast model.
When selected, the button confirms that Gemini is ‘skipping in-depth thinking’ before delivering a quicker response. Google says the tool is designed for general questions where speed is prioritised over detailed analysis.
The update coincides with changes to usage limits across subscription plans. AI Pro users now receive 300 Thinking prompts and 100 Pro prompts per day, while AI Ultra users get 1,500 Thinking prompts and 500 Pro prompts daily.
Free users also gain access to the revised limits, listed as ‘Basic access’ for both the Thinking and Pro models. Google has not indicated whether the Fast model will receive the Answer Now feature.
The rollout follows the recent launch of Gemini’s Personal Intelligence feature, which allows the chatbot to draw on Google services such as Gmail and Search history. Google says Answer Now will replace the existing Skip button and is now available on Android, iOS, and the web.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
South Korea has narrowed its race to develop a sovereign AI model, eliminating Naver and NCSoft from the government-backed competition. LG AI Research, SK Telecom, and Upstage now advance toward final selection by 2027.
The Ministry of Science and ICT emphasised that independent AI must be trained from scratch with initialised weights. Models reusing pre-trained results, even open source, do not meet this standard.
A wild-card round allows previously eliminated teams to re-enter the competition. Despite this option, major companies have declined, citing unclear benefits and high resource demands.
Industry observers warn that reduced participation could slow momentum for South Korea’s AI ambitions. The outcome is expected to shape the country’s approach to homegrown AI and technological independence.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
A vulnerability in Google Calendar allowed attackers to bypass privacy controls by embedding hidden instructions in standard calendar invitations. The issue exploited how Gemini interprets natural language when analysing user schedules.
Researchers at Miggo found that malicious prompts could be placed inside event descriptions. When Gemini scanned calendar data to answer routine queries, it unknowingly processed the embedded instructions.
The exploit used indirect prompt injection, a technique in which harmful commands are hidden within legitimate content. The AI model treated the text as trusted context rather than a potential threat.
In the proof-of-concept attack, Gemini was instructed to summarise a user’s private meetings and store the information in a new calendar event. The attacker could then access the data without alerting the victim.
Google confirmed the findings and deployed a fix after responsible disclosure. The case highlights growing security risks linked to how AI systems interpret natural language inputs.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
Lawmakers in the EU are moving closer to forcing technology companies to pay news publishers for the use of journalistic material in model training, according to a draft copyright report circulating in the European Parliament.
The text forms part of a broader effort to update copyright enforcement as automated content systems expand across media and information markets.
Compromise amendments also widen the scope beyond payment obligations, bringing AI-generated deepfakes and synthetic manipulation into sharper focus.
MEPs argue that existing legal tools fail to offer sufficient protection for publishers, journalists and citizens when automated systems reproduce or distort original reporting.
The report reflects growing concern that platform-driven content extraction undermines the sustainability of professional journalism. Lawmakers are increasingly framing compensation mechanisms as a corrective measure rather than as voluntary licensing or opaque commercial arrangements.
If adopted, the position of the Parliament would add further regulatory pressure on large technology firms already facing tighter scrutiny under the Digital Markets Act and related digital legislation, reinforcing Europe’s push to assert control over data use, content value and democratic safeguards.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Several major AI companies appear slow to meet EU transparency obligations, raising concerns over compliance with the AI Act.
Under the regulation, developers of large foundation models must disclose information about training data sources, allowing creators to assess whether copyrighted material has been used.
Such disclosures are intended to offer a minimal baseline of transparency, covering the use of public datasets, licensed material and scraped websites.
While open-source providers such as Hugging Face have already published detailed templates, leading commercial developers have so far provided only broad descriptions of data usage instead of specific sources.
Formal enforcement of the rules will not begin until later in the year, extending a grace period for companies that released models after August 2025.
The European Commission has indicated willingness to impose fines if necessary, although it continues to assess whether newer models fall under immediate obligations.
The issue is likely to become politically sensitive, as stricter enforcement could affect US-based technology firms and intensify transatlantic tensions over digital regulation.
Transparency under the AI Act may therefore test both regulatory resolve and international relations as implementation moves closer.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
A new report by Anthropic suggests fears that AI will replace jobs remain overstated, with current use showing AI supporting workers rather than eliminating roles.
Analysis of millions of anonymised conversations with the Claude assistant indicates technology is mainly used to assist with specific tasks rather than full job automation.
The research shows AI affects occupations unevenly, reshaping work depending on role and skill level. Higher-skilled tasks, particularly in software development, dominate use, while some roles automate simpler activities rather than core responsibilities.
Productivity gains remain limited when tasks grow more complex, as reliability declines and human correction becomes necessary.
Geographic differences also shape adoption. Wealthier countries tend to use AI more frequently for work and personal activities, while lower-income economies rely more heavily on AI for education. Such patterns reflect different stages of adoption instead of a uniform global transformation.
Anthropic argues that understanding how AI is used matters as much as measuring adoption rates. The report suggests future economic impact will depend on experimentation, regulation and the balance between automation and collaboration, rather than widespread job displacement.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
Pressure is growing on Keir Starmer after more than 60 Labour MPs called for a UK ban on social media use for under-16s, arguing that children’s online safety requires firmer regulation instead of voluntary platform measures.
The signatories span Labour’s internal divides, including senior parliamentarians and former frontbenchers, signalling broad concern over the impact of social media on young people’s well-being, education and mental health.
Supporters of the proposal point to Australia’s recently implemented ban as a model worth following, suggesting that early evidence could guide UK policy development rather than prolonged inaction.
Starmer is understood to favour a cautious approach, preferring to assess the Australian experience before endorsing legislation, as peers prepare to vote on related measures in the coming days.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!
