Anthropic and Accenture Commit $2 Billion to Embedded AI Safety Evaluation
Anthropic and Accenture have announced a partnership aimed at embedding teams within Anthropic's organization to evaluate the safety of advanced artificial intelligence (AI) models. The collaboration involves an investment commitment of at least $1 billion from each company over five years, with funds directed toward building capacity in AI safety evaluation. The initiative will leverage Accenture’s acquired AI subsidiary Faculty and focus on evaluating AI models through red-teaming exercises, alignment assessments, and testing existing model safeguards. This embedded approach allows evaluators to monitor how models are trained and deployed while assessing internal processes and verifying safety commitments. The companies note that standards for such evaluations remain undefined but emphasize the importance of this emerging field. Accenture’s CTO Marc Warner highlighted Faculty's mission to ensure AI is safe by design, citing its prior work with governments and healthcare systems like the UK NHS during the pandemic. Accenture CEO Julie Sweet emphasized the need for technical expertise combined with real-world understanding in ensuring safety.
Tickers: $ACN
Topics: Artificial Intelligence Partnership Safety Evaluation