Safety Research

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Aug 29

Lead enforcement against misuse of AI for cyberattacks and malware.

Anthropic · New York City, NY; San Francisco, CA +1 · Onsite · Jul 13

Build human review tooling that powers safety investigations and enforcement.

Anthropic · San Francisco, CA · Onsite · Jul 8

Build and run the Model Exploitation & Fraud team in Threat Intelligence.

Anthropic · San Francisco, CA | New York City, NY +1 · Onsite · Oct 2025

Build classifiers to detect misuse and anomalous behavior at scale.

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Jul 10

Review flagged activity to stop misuse for cyberattacks and malware operations.

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Jul 10

Build enforcement workflows tackling influence operations and disinformation.

Anthropic · San Francisco, CA | New York City, NY +1 · Onsite · Oct 9

Build safety and oversight mechanisms to monitor models and prevent misuse.

Anthropic · San Francisco, CA · Onsite · Aug 11

Prototype AI capabilities for real security team workflows on Claude Security.

Anthropic · San Francisco, CA | New York City, NY | Seattle, WA +1 · Onsite · Feb 24

Build multi-layered defenses to keep compute capacity fair and block bad actors.

Anthropic · Remote-Friendly (Travel-Required) | San Francisco, CA | Washington, DC +2 · Remote · Jan 14

Detect and disrupt misuse of AI for malicious cyber operations

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Aug 22

Own Anthropic's conventional weapons policy for how Claude can be used.

Anthropic · Remote-Friendly, United States; San Francisco, CA | Washington, DC +1 · Remote · Sep 1

Build and run evals measuring cyber capabilities and safeguard robustness in models.

Anthropic · New York City, NY; San Francisco, CA +1 · Onsite · Sep 15

Lead multi-month infrastructure projects from ambiguous ideas to production systems.

Anthropic · San Francisco, CA | New York City, NY +2 · Onsite · Oct 2025

Build agentic safety systems that help trust teams investigate abuse at model speed.

Anthropic · San Francisco, CA · Onsite · Jun 10

Lead review tooling that helps humans and Claude investigate potential harms.

Anthropic · San Francisco, CA · Onsite · Sep 2

Lead team building biosafety evaluations and classifiers for frontier models.

Anthropic · Remote-Friendly (Travel Required) | San Francisco, CA +1 · Remote · Jul 10

Adversarially test product surfaces to uncover vulnerabilities before attackers do.

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Remote · Mar 9

Run and monitor safety evaluations to ensure models meet policy standards.

Anthropic · San Francisco, CA | Seattle, WA | New York City, NY | Washington, DC +1 · Onsite · Sep 23

Own onboarding to offboarding for global human review workforce fighting misuse

Anthropic · Remote-Friendly (Travel-Required) | San Francisco, CA | Seattle, WA | New York City, NY +1 · Remote · Aug 24

Set safeguards approach for assessing product launches for safety risks.

Anthropic · New York City, NY · Onsite · Aug 14

Build data foundations powering safety detection, evaluation, and review systems.

Anthropic · London, UK · Onsite · Jan 21

Build foundational systems powering safety, oversight and intervention for AI.

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Sep 2

Lead child safety, mental health and abuse enforcement workflows at scale.

Anthropic · San Francisco, CA | New York City, NY | Washington, DC +1 · Onsite · Sep 13

Own misuse detection and enforcement across Bedrock, GCP, and Foundry surfaces.

Anthropic · San Francisco, CA | New York City, NY | Seattle, WA +1 · Onsite · Jul 10

Build real-time risk decisioning that scores transactions at authorization time.

Isomorphic Labs · Lausanne; London +1 · Onsite · Jul 3

Architect security for an AI-first platform powering next-gen drug discovery.

Waymo · Mountain View, CA, USA +1 · Onsite · Sep 14

Build models of human driving behavior like collision avoidance and interaction.

Biohub · New York, NY (Hybrid) +1 · Hybrid · Aug 12

Design safety evaluations for frontier AI capabilities in biology.

Biohub · Redwood City, CA (Hybrid) +1 · Hybrid · Jun 23

Design safety evaluations for frontier AI capabilities in biology.

ElevenLabs · London · Remote · Oct 12

Work as a Data Scientist focused on AI safety challenges.