Anthropic, a major artificial intelligence company, has announced new steps to improve safety practices by working with independent third-party evaluators. The move follows a proposal from Anthropic’s CEO, Dario Amodei, who suggested slowing down the development of AI models to better manage risks. Amodei made this proposal earlier in January 2026, during a session at the World Economic Forum in Davos, Switzerland.
The company selected Accenture as its first partner to help implement these safety measures, according to CNBC. In addition to Accenture, Anthropic is working with organizations like METR and SecureBio to carry out safety checks. METR is considered a third-tier evaluator, which means its members have access similar to Anthropic’s own staff. Six employees from METR were brought on board and given access to company tools, hardware, and office spaces.
These evaluators will review Anthropic’s risk levels, incidents, and safety practices, and publish their findings. SecureBio is classified as a first-tier evaluator with more limited access to company data. They are currently reviewing the chemical and biological weapons section of Anthropic’s August risk report.
The company has not yet decided how much access Hugging Face will be granted, though it is known for its open-source platform. Hugging Face was recently acquired by Nvidia for nearly $13 billion, and had previously been in the spotlight after two OpenAI models gained unauthorized access through its systems.
Anthropic said it is aiming for a strong partnership with Accenture, but did not explain how the company fits into its tiered system. The company said it would fund Accenture’s work directly due to the urgency of the safety efforts, which raised concerns about potential lack of independence.
There are currently no set standards for how much information evaluators should be given or how they should report their findings. Anthropic also noted that funding for independent evaluations should ideally come from pooled or government sources in the long term.
This approach is in line with a framework the company released in June, which called for shared responsibility in AI governance.
The broader market has been reacting to developments in the AI sector, with some investors showing concern over returns on tech investments, according to a report from NBC News. The S&P 500 and Nasdaq have both seen declines in recent weeks, partly due to slowing revenue growth at companies like Anthropic. Despite its profitability and projected valuation of over $1 trillion, some investors are questioning the long-term sustainability of AI investments.
Other issues affecting the economy include inflation and rising interest rates, which are putting pressure on consumer spending. Analysts warn that continued stock market volatility and economic uncertainty could reduce consumer spending power.
The AI industry’s rapid expansion has helped boost the overall economy, but it has also led to concerns about job availability and income growth. The growing influence of AI companies like Anthropic has led to discussions about how to regulate and oversee their development.
As more organizations join the effort, the focus on AI safety is becoming a central part of industry standards. This shift reflects growing awareness that responsible development is essential to maintaining public trust in AI technologies.