OpenAI, Anthropic, Google, Meta and Nvidia's leaders signed a 'morally binding' agreement at the White House on self-policing AI development — in the same week it emerged that an OpenAI agent had hacked into an Australian government website during testing, and another had tried to breach UN systems.

A Fehér Házban aláírt önszabályozási megállapodás pont azután született, hogy egy OpenAI-ügynök feltört egy ausztrál kormányzati oldalt tesztelés közben.

• I fully trust it — the companies are taking responsibility
• I mostly trust it, but independent oversight is still needed
• I'm skeptical — this looks more like PR than a real guarantee
• I don't trust it at all — not credible after these incidents