Anthropic says it has blocked multiple accounts that used the company’s artificial intelligence models in ways that could support the development of biological weapons.
In a report published Thursday, the AI company known for its Claude model said it considers biological misuse one of the “most serious risks” to artificial intelligence models.
The new report outlines five real-life case studies in which users “circumvented controls” that block users from specific regions and “engaged in other efforts to obfuscate the purpose of their research to evade our safeguards.” The examples include possible gain-of-function research and involve both infectious diseases, such as bird flu, and novel venoms and toxins.
I feel like it’s significantly easier to track the sale of certain equipment than trying to prevent the knowledge from getting around.
Given several glass jars, a pressure cooker, and some beef, you could make enough botulin to kill humanity within a few weeks.
The problem is delivering it.
US Air Force has a publicly available manual with instructions for the first part.