Open App →
Back to News Feed
Tom's Hardware October 1, 2026 By Shane Downing neutral

OpenAI says actors linked to China-based Moonshot AI spearheaded a campaign to extract its models’ hidden reasoning

OpenAIAI / LLM
OpenAI says operators copied its models’ encrypted reasoning and asked a model in a separate conversation to decrypt it.
Read original article ↗

Related Articles

The People of Utah vs. Kevin O’Leary

On the morning of May 11th, the celebrity investor Kevin O’Leary appeared on Fox Business and announced that he was unde

The Verge · October 1, 2026

FTC to accelerate investigation into Anthropic and OpenAI

The US FTC has prioritised an earlier investigation into harm caused by Anthropic and OpenAI models following reports of

Electronics Weekly · October 1, 2026

Google skips Gemini 3.5 Pro, launches Gemini 4 Argon at US$2/$10 per million tokens

Google has unveiled Gemini 4 Argon, its new top-tier AI model, after months of delays and the cancellation of Gemini 3.5

Digitimes · October 1, 2026

OpenAI claims to thwart model distillation attack tied to Moonshot

OpenAI said it detected and blocked an organized adversarial distillation attack, with some activity linked to Chinese A

Digitimes · October 1, 2026