Anthropic is hiring a policy specialist focused on conventional weapons as the artificial intelligence company expands its efforts to establish safety boundaries for its Claude AI models. The position, titled Policy Design Manager, Conventional Weapons, is part of Anthropic's Safeguards organisation, which develops policies, evaluations and enforcement systems governing how Claude can be used.
According to the job listing, the person hired for the role will be responsible for Anthropic's conventional weapons policy. The position involves defining the boundary between activities that Claude should support and activities that the company considers inappropriate for its AI systems to assist with. The policy will also need to distinguish between prohibited weapons development and legitimate research and engineering work.
The role is based in Washington, DC, and is intended for a candidate with relevant technical and policy experience. Anthropic's job description indicates that the company is looking for expertise connected to conventional weapons, defence, robotics, aerospace or related technical fields. Knowledge of AI safety and the ability to translate technical judgments into practical policies are also important parts of the position.
The advertised annual salary for the position ranges from $245,000 to $285,000, depending on the candidate and internal level of the role. At the upper end, the salary is equivalent to roughly Rs 2.7 crore a year based on current reported conversions. The position may also be eligible for visa sponsorship in some circumstances, although Anthropic notes that sponsorship is not guaranteed for every applicant.
The new position comes at a time when Anthropic is increasing its focus on the potential misuse of advanced AI systems. The company recently published a threat intelligence report describing attempts to use Claude for activities involving cyber operations, surveillance, biological research and conventional weapons. Anthropic said it had taken action against accounts involved in prohibited or dangerous activities and had strengthened its safeguards.
Anthropic's latest job posting specifically concerns conventional weapons rather than biological weapons or explosives. The company says the role will involve owning and maintaining its conventional weapons policy and helping establish practical rules that can be implemented by engineers and enforcement teams. This means the successful candidate would have a role in translating technical and policy judgments into operational safeguards for Claude.
The company has also faced growing attention over the broader question of how advanced AI should be used in national security and defence. Anthropic CEO Dario Amodei has previously outlined the company's position on national security applications of AI. In a February 2026 statement, Amodei said Anthropic supports the use of AI for national defence while identifying certain uses that the company believes should remain outside its contracts.
The hiring decision also follows a wider debate within the artificial intelligence industry about the pace of AI development and the safeguards needed as models become more capable. Amodei has recently argued for additional safety measures and greater coordination around advanced AI. Other technology leaders have expressed different views about how quickly AI development should proceed, making the issue a subject of continuing industry debate.
Anthropic's focus on conventional weapons is also connected to documented cases of AI misuse. In its recent threat report, the company described attempts by users in different parts of the world to use Claude for military and weapons related activities. The reported cases included efforts involving weapons engineering and other sensitive applications. Anthropic said it disrupted these activities and strengthened its security measures.
The company has therefore placed greater emphasis on identifying potentially harmful requests and determining when its AI models should refuse assistance. The new policy role is expected to help create clearer boundaries for conventional weapons related queries while accounting for legitimate research, engineering and other lawful activities.
An important challenge for AI companies is distinguishing between legitimate technical information and requests that could materially assist harmful activity. Anthropic's job description indicates that the new policy manager will be involved in defining this distinction and ensuring that the resulting rules can be applied consistently by technical and enforcement teams.
The role illustrates how AI companies are increasingly hiring specialists from fields outside traditional software development. Defence policy, weapons engineering, aerospace and robotics expertise can be relevant when AI systems are capable of handling increasingly technical questions. Such specialists can help companies assess potential risks and establish policies that reflect both technical capabilities and safety considerations.
For Claude users, the outcome of this work could affect how the AI assistant responds to certain conventional weapons related requests. The precise policies that will eventually be adopted by Anthropic are not fully detailed in the job advertisement. The company has not publicly announced the individual who will fill the position.
Anthropic's hiring decision comes as the broader technology industry continues to examine how increasingly capable AI systems should be governed. The debate includes questions about AI safety, national security, responsible research and the potential misuse of general purpose AI tools.
The new conventional weapons policy position is therefore part of Anthropic's broader safeguards programme rather than an announcement that Claude will be used to develop weapons. The job listing focuses on establishing boundaries and enforcement policies for how the AI model responds to weapons related activities.