Anthropic, the cutting-edge US artificial intelligence powerhouse, is aggressively recruiting a chemical weapons and high-yield explosives specialist to guard against the catastrophic misuse of its AI technology.
Securing AI Against Dangerous Weaponization
The company confronts a chilling concern: its sophisticated AI tools could potentially instruct malicious actors on creating chemical, radiological, or explosive weapons. To counter this, Anthropic demands an expert with over five years of experience in chemical weapons and explosives defense, coupled with deep knowledge of radiological dispersal devices—commonly known as dirty bombs.

In a detailed LinkedIn job posting, the firm emphasizes the critical nature of this role, which mirrors other sensitive positions it has recently established. The goal is clear — to build robust guardrails that prevent dangerous instructions from ever leaving their systems.
Industry-Wide Moves to Tackle AI Weaponization Risks
Anthropic is not alone in this defensive strategy. OpenAI, creator of ChatGPT, has similarly advertised a high-profile researcher role focused on biological and chemical risks, offering a salary nearing $455,000 (£335,000) — nearly twice Anthropic’s offer. This highlights the escalating urgency across AI firms to manage the dark side of AI capabilities.
Experts Warn of Dangerous Trade-Offs
Yet, some authorities voice serious concerns. Dr. Stephanie Hare, a tech researcher and co-host of the BBC’s AI Decoded, questions the wisdom of involving AI in sensitive weapons knowledge. “Is it ever safe to use AI systems to handle sensitive chemicals and explosives information, including dirty bombs and other radiological weapons?” she asks. She stresses the absence of any international treaties or regulations governing AI’s use in weapons development, warning that these developments unfold largely in the shadows.

Geopolitical Stakes and AI’s Existential Threats
The AI sector has persistently sounded alarms about existential threats posed by its technology, yet rapid innovation continues unabated. This urgency intensifies amid US government involvement, particularly as military operations escalate in Iran and Venezuela.
Anthropic recently escalated tensions by suing the US Department of Defense after being labeled a supply chain security risk. This designation stemmed from Anthropic’s principled refusal to allow its AI systems’ deployment in fully autonomous weapons or mass surveillance targeting American citizens.
Dario Amodei, Anthropic’s co-founder, declared in February that the technology remains immature for such applications and should be barred from them. However, the White House has asserted that military operations will not be constrained by tech companies’ ethical boundaries.
This risk classification places Anthropic alongside China’s telecom giant Huawei, which faces similar blacklisting over unrelated national security concerns.
Contrasting Government Relations Among AI Titans
OpenAI endorses Anthropic’s cautious stance but has taken a divergent path, negotiating a government contract that has yet to commence. Meanwhile, Anthropic’s AI assistant, Claude, remains actively integrated into platforms supplied by Palantir and continues deployment in the ongoing US-Israel military context involving Iran.
Anthropic’s Legal Battle and Industry Backing
Amid this fraught landscape, Anthropic garners support from major tech players as it challenges the Trump administration’s security risk designation. Simultaneously, AI safety advocates raise alarms about the global peril posed by unchecked AI weaponization, with some departing the field entirely to pursue alternative vocations, including poetry.









