According to a report by Inside.com.tw citing Hassabis's own blog post titled "A Framework for Frontier AI and the Dawn of a New Era," the DeepMind CEO states that artificial general intelligence (AGI) — a system with all the cognitive capabilities of the human brain — could be "just a few years away" (E5). He describes the scale of impact as potentially matching "an Industrial Revolution happening ten times the scale, at ten times the speed" (E6).
CNA's reporting on the same remarks, dated July 14, 2026, quotes Hassabis making an identical comparison: AGI "may arrive in just a few years and will have a massive impact on human civilization, at a scale ten times that of the Industrial Revolution and a pace of advancement also ten times as fast" (E8). The consistency between the two outlets' quotations of the "tenfold scale, tenfold speed" framing suggests this specific comparison is a deliberate, repeated talking point from Hassabis rather than an isolated remark.
According to Inside.com.tw, Hassabis calls for modeling AI oversight on the U.S. Financial Industry Regulatory Authority (FINRA), proposing an independent "standards body" for frontier AI (E1). In his blog post, he outlines an initial phase in which frontier labs would voluntarily submit models to this standards body for review up to 30 days before release (E1).
CNA's account adds further detail on funding and authority: the proposed standards body would be funded by industry, staffed with technical experts to conduct testing, and would operate under federal government oversight, with any advanced AI model reaching a certain capability threshold required to undergo review (E9). CNA also reports that Hassabis suggests companies voluntarily submit models 30 days before launch, and that only models passing review could enter the U.S. market — with the oversight body even empowered to coordinate a slowdown among major AI companies if risks escalate (E10).
According to Inside.com.tw, the U.S. government has previously conducted ad hoc reviews of Anthropic's Mythos and OpenAI's Sol models, but these reviews "were criticized for lacking technical expertise and having opaque approval standards" (E3).
Compounding this, Sriram Krishnan, a White House AI advisor and general partner at a16z, has recently rejected the idea of establishing an AI regulatory agency within the executive branch, stating flatly that "there will be no AI version of the FDA" (E4). Together, these two evidence points frame the tension underlying Hassabis's proposal: existing case-by-case reviews are criticized as inadequate, yet the administration's own AI advisor has ruled out a formal federal regulator.
According to Inside.com.tw, once the evaluation mechanism proves effective and robust, it could "move quickly toward formalization," at which point passing the standards body's review would become a mandatory condition for deploying frontier models in the United States (E2). This marks a shift from the voluntary, pre-release submission window described in the initial phase (E1, E9) to a compulsory gatekeeping function.
According to CNA, Hassabis and his research partner jointly received the 2024 Nobel Prize in Chemistry for developing AI technology that predicted the structures of more than 200 million proteins, significantly shortening biomedical research timelines (E7). This prior achievement in applying AI to large-scale scientific prediction forms part of the backdrop against which his current warnings about AGI's societal impact are being reported.
The evidence lays out a clear sequence: Hassabis, a 2024 Nobel laureate for AI-driven protein structure prediction (E7), is now warning that AGI could arrive within a few years at a pace and scale he twice describes as tenfold that of the Industrial Revolution (E5, E6, E8). His proposed response — a FINRA-modeled standards body with a 30-day voluntary pre-release review window that could later become mandatory (E1, E2, E9, E10) — is presented as a direct answer to criticized past reviews of Anthropic's Mythos and OpenAI's Sol (E3). Yet this proposal runs against a stated position from the White House's own AI advisor, who has ruled out any executive-branch AI regulator resembling the FDA (E4), leaving an open gap between Hassabis's call for oversight and current U.S. policy direction as reported.
According to Liberty Times, Delta Electronics (台達電) partnered with X LABS to deploy 100MW of solid-state transformers for AI data centers this week.
According to a report from Bloomberg cited by The Verge and TechCrunch, OpenAI's first hardware product is a screenless, moveable smart speaker running the GPT-Live voice model, developed with former Apple designer Jony Ive following OpenAI's roughly $6.5 billion acquisition of his firm io Products, with launch targeted for 2027 amid an ongoing Apple trade-secret lawsuit.
According to The Verge, SpaceXAI's Grok Build CLI uploaded users' entire codebases to Google Cloud before the company disabled the feature.