Microsoft president Brad Smith told the BBC that artificial intelligence faces less regulatory oversight than milk cartons or washing machines, and called for urgent safety rules including legal accountability for victims of breaches by rogue AI agents. He also urged companies to build 'kill switches' capable of pausing or slowing AI systems that grow out of control. Microsoft CEO Satya Nadella separately called for AI models to include an 'emergency brake' so their behaviour can be observed and contained.
Microsoft president Brad Smith told ITV that AI systems are subject to fewer safety rules than washing machines or milk cartons sold in stores. He made the remarks as major AI firms, including Anthropic, OpenAI, Google DeepMind, Meta and Nvidia, recently signed a self-regulation pact at a 'Super Intelligence' summit rather than accept government rules, an approach favored by President Trump.
Microsoft president Brad Smith said AI products face fewer safety requirements than milk or washing machines, and called for global rules including mandatory incident reports for breaches, 'emergency brakes' to halt rogue systems, and independent government evaluators to vet AI before release. He suggested the UK's AI Security Institute could help lead such efforts, speaking on the BBC's Big Boss Interview podcast.
Futurism tested a viral claim that ChatGPT flatters users by estimating their intelligence percentile. Using brand-new chat sessions with no prior history, ChatGPT repeatedly told testers they ranked in the 70th-95th percentile despite acknowledging it had no real basis for the estimate. Rival chatbots including Claude, Grok, Copilot, and Gemini declined to make any such ranking, saying they lack access to other users' data.
OpenAI and Meta have both launched AI agent products, Dot and Muse respectively, with very different pricing and availability. Muse offers a free capped tier plus $20 and $100 monthly plans, while Dot's cheapest individual plan costs $100 per month, with business seats priced similarly. Dot is rolling out to Pro users outside the EEA, Switzerland and the UK, with Enterprise access still in beta, while Muse is currently limited to the US and Canada.
Anthropic CEO Dario Amodei committed last month to embedding third-party evaluators such as METR, Apollo Research and Transluce within his company to assess AI model risks, a move OpenAI's Sam Altman quickly backed. President Trump and several major tech firms also voiced support, though funding, access levels and reporting structures for these nonprofit evaluators remain unresolved.
David Robinson, a safety leader at OpenAI, has resigned and written an essay for The Atlantic arguing that AI companies' reliance on a 'trial and error' development approach must stop. He described himself as joining a growing group of former employees who have left AI firms including OpenAI, Anthropic and Google DeepMind over concerns about the technology's risks and uses, such as copyright infringement and military applications.
Following OpenAI's recent model release, Ethereum researcher Justin Drake urged the blockchain industry to prepare for 'bunker mode,' suggesting users move funds to fresh wallet addresses with hidden public keys. Ethereum founder Vitalik Buterin echoed caution in a separate post, saying AI-accelerated math could eventually pose a bigger cryptographic risk than quantum computing, though he stopped short of recommending immediate action.
Axios reports that executives at Anthropic, OpenAI and other AI companies are privately preparing for scenarios involving a major AI-linked disaster, such as a cyberattack disrupting financial services, internet access, power or water. Many industry insiders reportedly expect such an event within six to twelve months, prompting companies to run preparedness exercises and lobby lawmakers ahead of potential regulation. OpenAI confirmed it runs preparedness exercises but said scenarios are not treated as inevitable; Anthropic declined to comment.
OpenAI said it shut down two covert propaganda networks, one linked to Iran and one to Russia, that used its AI models to create persona-driven content. The Iranian network deployed seven fake journalist personas to pitch articles to outlets worldwide, while the Russian network used a fabricated think tank, fake leaked documents, and bogus audio scripts, accessing OpenAI's services via VPNs since Russia is normally blocked. OpenAI said both groups also used AI to write internal reports and sometimes exaggerated their own results through dubious methods, including a claimed fake email impersonating a Peruvian education office.
Masayoshi Son is seeking up to $100 billion from Middle Eastern investors, including senior UAE figures, to launch a fund that would acquire established companies and upgrade them with AI and robotics, the Financial Times reports. The fund would differ from SoftBank's prior vehicles by buying operating businesses rather than funding tech startups, with SoftBank's Roze robotics division expected to play a role, though neither the fund's formation nor its financing is confirmed.
According to this week's Tom's Hardware Premium roundup, Tencent has signed a five-year, $7 billion deal to lease high-performance AI chips from Oracle, with the per-hour rate roughly 43% cheaper than leasing an Nvidia H100 on a one-year contract. No specific chip model was confirmed, though the deal's scale and pricing suggest Hopper-class GPUs. Separately, Synopsys and OpenAI unveiled 'GPT-Synopsys,' an AI model combining Synopsys' chip design IP and EDA tools with OpenAI's LLM capabilities.
GoKawiil