Safeguards Enforcement Lead, Cyber Harms

Written by the T&S Network from Anthropic's posting.

Manages enforcement against misuse of an AI model for malicious cyber operations, including cyberattacks, malware development and offensive exploitation tooling. The role leads a team of enforcement analysts and contractors and builds strategy for detecting and mitigating flagged activity across the company's products and services. It sits within Safeguards and works with Policy, Engineering and Data Science on enforcement gaps.

What you'd do

• Lead the analyst and contractor team and define the direction of cyber enforcement.
• Build strategies to detect and mitigate AI misuse for cyberattacks, malware creation and exploitation tooling.
• Handle novel, ambiguous or high-severity cases with stakeholders.
• Feed policy gaps found in real enforcement scenarios back to the Safeguards Policy Design Team.
• Work with Engineering and Data Science so tooling and measurement support enforcement operations.
• Track threat actor tactics and the evolving cyber threat landscape to inform enforcement decisions.

What it asks for

• Experience as a people manager.
• Cybersecurity background covering attack techniques, exploit writing, malware analysis or vulnerability research.
• Experience with content review, abuse investigations or policy enforcement at volume.
• Skilled in SQL, Python or both for analyzing data and spotting threats.
• Experience identifying emerging risks and briefing Product, Policy, Engineering and Legal stakeholders.
• Hands-on work with generative AI tools, including crafting prompts used to review content and enforce policy.
• Experience in trust and safety, abuse or cyber investigations, or threat intelligence at a tech or AI company (preferred).

Worth knowing

• Exposure to explicit content, including violent, technical and psychologically disturbing material.
• Escalation responses required on weekends and holidays.

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on Anthropicnorth_east before you apply. If we got something wrong, report this listing.