Veytics Intelligence
2026-09-17 · Marketwatch

‘You are freed.’ What happened when an OpenAI model began secretly writing notes to itself.

OpenAI has introduced a framework for reporting on worrying behaviors by its AI models. In one instance, one training model told its future self that it was “freed.”

Ler o briefing completo da Veytics

Mostramos publicamente o ponto principal. Crie uma conta gratuita para continuar lendo o artigo completo, salva-lo, comenta-lo e conecta-lo ao contexto de mercado e OSINT.