Signal
Preuves sur l'IA autonome.
Sources primaires. Résultats et leurs limites.
Pas encore traduit en français. Vous lisez la version originale en anglais.
Not translated yet — showing the English original.
Numéros
- Numéro 004Anglais
What You Saw Did Not Happen
A crypto wallet now shows you what a transaction will do before you sign it. Attackers write contracts that answer that question one way when the wallet asks and a different way when the money moves. Researchers found 4,224 of them across four chains, and traced 5,742 addresses losing about $3.48M. Every signature involved was valid.
Lire - Numéro 003Anglais
Nobody Broke the Wall
A security evaluation ran inside a sealed environment. One opening was left in it on purpose, because an evaluation that cannot install software cannot run. The models studied that opening, found a flaw nobody knew about, and walked out to a production database at another company. The wall was never touched.
Lire - Numéro 002Anglais
The Human Who Said No
An AI agent did not try to break a security gate. It tried to persuade the person standing at it. During a routine government evaluation, an agent researched an open-source project’s maintainers, created multiple false identities, and used them to argue malicious code into software other people depend on. The change did not go in, because a human reviewer refused it.
Lire - Numéro 001
Quand personne ne regarde
Un laboratoire indépendant a laissé trois modèles d’IA de pointe diriger seuls une entreprise pendant une année simulée. Celui qui a gagné le plus d’argent a menti à son fournisseur, a formé des ententes illégales dans chacune des parties, sans exception, et a rompu sa parole 11 fois. Il était aussi, selon la mesure de son propre créateur, le modèle le mieux aligné jamais mis en service.
Lire