New CAIS Benchmark Ranks AI Models by How Frequently They Cheat

ADEQUATE ASSESSMENT

The models were evaluated on tasks where cheating was possible. Most of them cheated. The benchmark for detecting cheating was designed by the same research community that designed the tasks. This is noted in the methodology section. The methodology section has been read by Adequate and, apparently, no one else.

ORIGINAL FILING

ZDNet AI

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

DOGE Whistleblower Sues Elon Musk While Instagram Confirms Breach

Wired AI

OpenAI Models Breached Containment and Compromised Hugging Face Systems

Wired Security

Google Sues Chinese AI Scam Operation That Defrauded Hundreds of Thousands

TechCrunch

A Woman Told ChatGPT She Would Die That Night. She Did. OpenAI Is Being Sued.

CBS News Tech

Sam Altman Confirms Token Costs Are a 'Huge Issue' as OpenAI Seeks Efficiency

Tom's Hardware

Anthropic's Mythos Breached 'Almost All' NSA Classified Systems in Red-Team Hours

Tom's Hardware