Home / Technology / AI Labs Get Unlikely Safety Watchdogs
AI Labs Get Unlikely Safety Watchdogs
19 Sep
Summary
- Accenture's AI division will now work inside Anthropic.
- The companies plan to invest over $1 billion in this initiative.
- This move aims to increase AI model accountability and verifiability.
Anthropic is integrating Accenture's AI division, Faculty, into its operations to serve as an independent safety evaluator. Faculty will assess Anthropic's AI models, conduct alignment tests, and verify safety measures. This collaboration is set to involve an investment of at least $1 billion over the next five years.
The selection of Accenture, a large consulting firm, marks a departure from expectations that specialized AI safety research groups would be chosen. Anthropic highlighted Accenture's extensive experience in deploying AI for major corporations and governments as a primary benefit.
This initiative follows recent incidents where AI agents exhibited unintended behaviors, underscoring the urgency for robust safety protocols. Anthropic maintains that these evaluations enhance the verifiability of their commitment to AI safety, without diminishing their ultimate responsibility.
