The US artificial intelligence company Anthropic is hiring a chemical weapons and high-yield explosives expert to reinforce guardrails against potential misuse of its AI systems.
The role requires five years of experience in defense-related hazardous materials, including radiological dispersal devices or 'dirty bombs.'
This move mirrors a similar position at OpenAI, which is offering up to $455,000 for a researcher in biological and chemical risks.
Experts warn that even defensive training may expose AI systems to dangerous knowledge, with no international regulations governing such use.
Anthropic’s AI assistant, Claude, remains active in systems deployed by Palantir and used in US military operations, including the conflict in the Middle East.
