AI model maker Anthropic announced it will grant six researchers from the nonprofit METR employee-level access to independently evaluate its AI models, safety practices, and risk management. Anthropic has waived its right to block METR from publishing evaluation results, but can request edits for predefined categories such as customer data, attorney-client privileged information, or model architecture details. Anthropic likened the arrangement to bank regulators embedding examiners within large banks. This move fulfills a promise made by Anthropic CEO Dario Amodei in September to slow down AI development through embedded external evaluators.