Live Updates > Live Update Details
2026-09-19 04:54:09
On September 18, AI company Anthropic announced a collaboration with professional services firm Accenture. Anthropic will establish an embedded evaluation team within Accenture to conduct red-team testing, security mechanism verification, and alignment evaluation of cutting-edge AI models, ensuring that technological development aligns with human expectations. Both companies plan to invest at least $1 billion over the next five years to build this capability system. Accenture's evaluators will have the same internal access as Anthropic employees, enabling them to monitor the model training iteration process in real time and observe the entire decision-making process. Anthropic stated that the purpose of this collaboration is to verify the implementation of its security commitments, identify previously undiscovered security blind spots, and provide the public with more transparent risk assessment results.
