Tech Shots
AI news flashes
עברית
Deals

$2B for AI safety — Accenture embeds inside Anthropic

Anthropic and Accenture each committed at least $1 billion — about $2 billion combined over five years — to frontier model evaluation and safety, with Accenture’s Faculty as Anthropic’s first “embedded evaluator.” Per Reuters and TechCrunch, Faculty will evaluate and red-team models, run alignment assessments and test safeguards, with employee-comparable access to verify safety commitments and report incidents. Anthropic says more evaluators will follow, including talks with METR and other nonprofits.

Source: Reuters