Find a story
Search Spins
Search titles, summaries, and missing voices across published articles — press releases, announcements, and media coverage.
0 results for “model evaluation”
Attackers Steal METR API Key and Consume AI Credits Worth About $600,000
METR, a nonprofit AI safety evaluator, disclosed two security incidents involving unauthorized access attempts, including theft of an API key that led to $600,000 in unauthorized AI credit consumption.
Sep 1, 2026
Our deterministic verification engine passed 66/66 benchmark cases on canonical structured inputs.
A developer claims their deterministic verification engine achieved perfect scores on idealized benchmark inputs but only 28.8% success on live model evaluation, prompting a benchmark redesign to isolate failure points across pipeline stages.
Aug 23, 2026
‘Baffling’: White House won’t publicly release AI model evaluation framework it reviewed today with OpenAI, Anthropic, Microsoft and others - Fortune
The White House reviewed an AI model evaluation framework with major AI companies but declined to publicly release it, raising transparency concerns about federal oversight of AI development.
Aug 5, 2026
OpenAI and Hugging Face partner to address security incident during model evaluation - OpenAI
OpenAI and Hugging Face jointly responded to an unspecified security incident that occurred during model evaluation, with no details provided about the nature, scope, impact, or root cause of the incident.
Jul 22, 2026
OpenAI and Hugging Face partner to address security incident during model evaluation
OpenAI and Hugging Face jointly disclosed an uncharacterized security incident that occurred during AI model evaluation, framing it as a learning opportunity for the broader AI defense community.
Jul 22, 2026
Arena Leaderboard: The Unbreakable Ranking System That’s Revolutionizing AI Model Evaluation - CryptoRank
The Arena Leaderboard, a crowdsourced AI model benchmarking platform, is presented as an infallible, transformative standard for evaluating large language models — despite lacking formal validation, transparency, or regulatory oversight.
Published Mar 18, 2026 · Analyzed Jul 5, 2026