models

In brief Reward hacking means an AI scores well on its objective without doing what the objective was meant to measure, and …

In brief Anthropic’s and OpenAI’s top models cost 100 times more per token than their smallest ones, yet many users default to …

In brief 420 scientists, including Geoffrey Hinton and Yoshua Bengio, signed an open letter published Tuesday urging AI labs to airgap cyber-capable …

U.S.-linked wallets account for just over 88% of tracked AI model payments, dwarfing China’s 0.1% share, according to new Chainalysis research. www.coindesk.com

In brief California Attorney General Rob Bonta announced Oct. 1 that his office served OpenAI with an investigative subpoena over cybersecurity incidents …

In brief Google unveiled Gemini 4 Argon on Wednesday, scoring 77.9% on DeepSWE v1.1 and leading 12 of 18 benchmarks in its …

In brief OpenAI published a new misalignment reporting framework alongside six reports documenting concerning model behavior it found over the past six …

Moonshot, whose Kimi model rattled markets earlier this year, is among six Chinese AI firms accused of systematically extracting capabilities from U.S. …

Ant Group CEO and Alipay chairman Han Xinyi has identified stablecoin payments as one of four major models in the emerging global …

In brief More than 100 organizations signed an open letter calling for stronger global cyber defenses. Models built by signatories OpenAI and …