Search
In a bid to uphold its safety pledges and alleviate external apprehensions regarding the potentially catastrophic risks of AI technology spiraling out of control, Anthropic has recently declared that it will provide external evaluation entity METR with access privileges akin to those enjoyed by its own employees. This will empower METR to evaluate the potential hazards associated with the Claude model. Presently, the METR team has set up operations within Anthropic's premises, with unrestricted access to hardware facilities and internal systems. This strategic initiative is designed to bolster safety safeguards through independent third-party evaluations. Concurrently, Anthropic is partnering with Accenture to develop a safety protection framework valued at over $1 billion and has engaged SecureBio to scrutinize risks pertaining to chemical and biological weaponry.
-
C114 Communication Network -
Communication Home
