AI: Washington Upholds Voluntary Assessment of Models

Estonia reports increased Russian hacking capabilities due to AI. In Washington, Sean Cairncross defends the potential of this technology for protection while dismissing a mandatory assessment of new advanced models.
Kristen Michal describes an escalation of attacks
Russian hacking capabilities have advanced thanks to AI over the past year, claims Kristen Michal, head of the Estonian government. According to this official, attacks targeting her country's highly connected administration have become more frequent. Michal attributes these to both private actors and to Russia, Iran, or North Korea.
When asked whether a cyberattack could constitute an act of war, Michal believes that some could fall under NATO's Article 5. This classification primarily depends, in her view, on the consequences of the attack. Present alongside her on stage, Sean Cairncross expressed confidence in the defensive capabilities of AI, which he considers to be previously inconceivable.
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
The White House prioritizes laboratory cooperation
On Monday, during Silicon Valley & The World, an event organized by Semafor, the White House cyber chief indicated that the Trump administration did not plan to make its voluntary assessment framework public or impose its application. This framework was created by executive order in June. Cairncross justifies the lack of full publication by the desire to maintain open exchanges with companies but is convinced that Washington will soon clarify the concrete modalities. He praises the cooperation of leading laboratories, the quality of their interactions with the administration, and the submission of their new models for evaluation.
According to Cairncross, a strict framework would not keep pace with the necessary speed. The U.S. administration applies lighter constraints to the AI industry than those in Europe or under the Biden presidency. It has also initiated measures aimed at preventing states from intervening in this area.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.