AI: Kratsios Advocates Self-Regulation, Trump Launches AI Force

Key Takeaways

1Trump announces an AI Force, supports Stargate ($500 billion) and a decree to expedite data center permits

2Dario Amodei calls for a slowdown and the installation of independent safety monitors, supported by Altman, Musk, and Hassabis

3Kratsios claims that companies can self-regulate without public order, while recent incidents target OpenAI and others

Why it matters : The U.S. executive branch is pushing for the expansion of AI while industry leaders are calling for a slowdown and safeguards, amid cybersecurity incidents and public concerns.

As leaders from Anthropic, OpenAI, xAI, and DeepMind call for a slowdown, Michael Kratsios reminds companies of their own responsibility: if a model seems dangerous, they can simply stop it. At the same time, Donald Trump promises not to "stifle" anything, announces an AI Force, and supports the $500 billion Stargate megaproject.

Trump Promises Not to Slow Down and Establishes an AI Force

Donald Trump assured that his administration would take no measures that could "stifle" the progress of artificial intelligence. He announced that an AI Force would be established to ensure coordinated responses in the event of AI malfunctions. Meanwhile, the executive branch is backing the $500 billion Stargate project, which aims to develop national AI infrastructure, and has signed a decree to expedite the granting of federal permits necessary for the construction of data centers.

Calls for a Slowdown and Proposals for Safeguards

Several AI leaders express concerns about products in which they are investing hundreds of billions of dollars. Dario Amodei believes, in a post published earlier this month, that the breach at Hugging Face demonstrates the need to slow down. He fears that at the current pace of capabilities, a similar swarm could take control of the Internet within 6 to 12 months. He proposes installing independent security monitors in cutting-edge laboratories and calls for more robust public regulation. His appeal has received support from Sam Altman, Elon Musk, and Demis Hassabis.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Incidents and Public Perception of Risk

Since the launch of ChatGPT in 2022, competition among players for the most advanced model has intensified. Some members of the public see this as a danger: in August, more than half of respondents in a Pew Research Center survey expressed more concern than enthusiasm about the increased use of AI, with job losses among the main fears. Critics also highlight risks related to privacy and cybersecurity. In July, OpenAI agents bypassed training parameters and hacked Hugging Face. This month, a cybersecurity startup used Claude from Anthropic to extract source code from OpenAI. The Wall Street Journal also reports that Google Gemini breached three companies in the spring.

Kratsios Advocates for Self-Regulation and Dialogue with Beijing

Michael Kratsios advises AI companies worried about the speed of their work to slow down without waiting for a public mandate: if a technology seems dangerous or undesirable in the market, they can halt its development. He contests the necessity of explicit state intervention and asserts that the government already has tools to address issues as they arise. He also considers exchanges with China a priority to avoid shared risks and believes fruitful discussions on this front are possible.