US AI company Anthropic says it will allow more users to access its leading-edge Claude Mythos model.
Now more companies and organizations can apply to use the AI agent for so-called defensive work in their computer systems.
These include universities, government entities and infrastructure-related firms such as power companies.
Access tiers are to be separated according to the intended use of the AI model.
For the highest tier, the company says it is screening applicants in cooperation with the US government.
The AI developer has been limiting access over concern the model’s advanced capabilities could allow it to be used in cyberattacks.
As of April this year, access to Claude Mythos was limited to about 50 companies and organizations. That expanded in June to about 200, including Japanese financial institutions.
Source: www3.nhk.or.jp
