OpenAI, Microsoft and Anthropic sign Paris AI safety pact as critics flag weak enforcement
Ten AI firms, including OpenAI and Microsoft, signed the Paris Call for AI Safety and Security, agreeing to new risk-sharing measures. Critics say the agreement lacks enforceable standards. Regulated firms should watch for evolving oversight and compliance expectations.
OpenAI and Anthropic must comply with New York’s new AI safety law by 2027
New York’s RAISE Act now requires large AI developers like OpenAI and Anthropic to register and publish safety frameworks. Senior leaders in regulated industries face new compliance and reporting risks as enforcement ramps up. Penalties reach $3 million per violation.
UK banks cut AI exposure as Bank of England flags sector risks
UK banks are reducing their exposure to artificial intelligence, the Bank of England’s Prudential Regulation Authority chief said. This move comes amid rising concerns about volatility in AI markets. Senior leaders should watch for new regulatory guidance on AI and operational resilience.
OpenAI sets limits on ad revenue ambitions to protect user privacy
OpenAI’s monetisation chief says privacy rules restrict ad partner requests and distances from the $100 billion ad revenue figure. Senior leaders in regulated sectors should note OpenAI’s stance on user trust over aggressive monetisation. Watch for evolving privacy standards in AI-driven advertising.
Hospitals lose $132B despite AI and automation investments
Hospitals have written off $132 billion as bad debt, according to Healthcare Dive. This comes despite large investments in AI and automation aimed at improving financial results. Senior healthcare leaders face pressure to show returns on digital spend.
Palantir pressed UK minister to expand AI access to NHS GP records, raising data trust issues
Palantir lobbied a UK health minister to include sensitive GP records in its NHS data platform, despite government promises to exclude them. This move raises concerns for healthcare leaders about data privacy, AI governance and patient trust. Watch for government decisions on contract scope and future suppliers.
US AI safety centre moves under White House as experts warn of political risk
The US government’s AI safety division now reports to the White House, according to two former officials. Critics say this could politicise oversight of AI risks, which matter for regulated sectors. Executives should monitor for shifts in AI governance and regulatory approach.
Australia plans criminal law changes and mandatory AI breach reporting after OpenAI incident
Australia will introduce mandatory reporting and consider criminal penalties for AI firms after an OpenAI-linked Medicare data breach. Senior leaders in regulated sectors face stricter oversight for AI risk and incident response. Watch for new AI safety rules and enforcement timelines.
Berkshire Hathaway spends $9.6bn on buybacks as insurance earnings fall
Berkshire Hathaway bought back $9.6 billion in shares after a 43% drop in insurance underwriting profits. The move signals a shift in capital allocation as operating income fell 5.8%. Insurers and finance leaders should watch for further buybacks and AI bets.
AI-enabled cyber tools used in new attacks on banks, says Cybersecurity Dive
A recent campaign used Artex and other pen-testing tools to automate attacks on banks, according to Cybersecurity Dive. This shows cybercriminals are adopting AI tools for financial gain. Senior leaders should review cyber risk controls as automation rises.
Ogilvy and CAA sign deal to use AI clones of celebrities in ads, cutting costs by up to 75%
Ogilvy and Creative Artists Agency have agreed a partnership to create AI-generated ads using digital clones of actors and athletes. The move could cut campaign costs by up to 75% and speed up production, but raises questions about trust and authenticity for brand leaders.
Google’s AI visibility changes test UK-mandated search opt-out for publishers
Google has rolled out a global search opt-out for publishers, following a UK regulator’s order, but bundled controls and missing click data complicate decisions. Senior leaders must weigh visibility against control and revenue. Watch for Google’s next move on click data.
Tencent AI chief warns of biological risks from open AI models
Tencent’s AI chief said AI models could soon cause harm in biology, not just cybersecurity. This raises new risk management questions for pharma, healthcare and regulated sectors. Executives should follow the debate on open versus closed AI models.
AI-generated complaints swamp UK public bodies and employers
UK public bodies and businesses are facing a surge of AI-generated complaints and legal challenges, with employment tribunal claims up 28%. This trend increases operational pressure and legal risk for regulated firms. Leaders may need to review how they handle incoming correspondence.
Anthropic, Google, Microsoft and OpenAI back AI training opt-out token for publishers
Anthropic, Google, Microsoft and OpenAI say they will support a new AI training opt-out policy token. The move gives publishers a standard way to signal if their content can be used for AI training, but enforcement remains weak. Regulated firms face similar data governance issues.
AI exposes IT flaws as $6bn Binance hack and fake nuclear alerts hit critical systems
AI is revealing major IT vulnerabilities, with incidents including a $6.3bn Binance heist and fake nuclear alerts, according to the UK’s National Cyber Security Centre. Banks, insurers and healthcare providers face growing risks to critical infrastructure. Watch for regulatory responses and new security standards.
US Justice Department forms Super Intelligence Task Force to target AI crimes
The US Justice Department has set up a Super Intelligence Task Force to lead on AI crimes and lawsuits. This move signals more scrutiny for firms using AI in regulated sectors. Executives should expect new compliance demands and enforcement priorities.
AI-driven drug research disappoints again in pharma sector
A promising AI technology aimed at accelerating drug research has failed to deliver, according to STAT. This setback raises questions about the reliability of AI in regulated biotech and pharma industries. Firms should review AI risk and validation processes.
US charges three over $129m AI voice fraud targeting executives
US prosecutors have charged three men with using AI to impersonate CEOs in a $129m fraud. The scheme exploited deepfake technology to trick employees at banks, fintechs and hospitals into wiring funds. Senior leaders face growing risk from AI-driven scams and must review controls.
Microsoft ends Oracle AI cloud deal as $80B contract review hits suppliers
Microsoft is ending its 2023 deal to rent AI computing from Oracle, declining renewal as it reassesses over $80B in outside cloud contracts. This shift affects AI infrastructure suppliers and signals cost and technology pressures. Watch for further supplier changes.