ML/Research Engineer, Safeguards
San Francisco, CA | New York City, NYFull-TimeMid-levelSoftware Engineering
About Anthropic
- Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. We want AI to be safe and beneficial for our users and for society as a whole. Our team is a quickly growing group of committed researchers, engineers, policy experts, and business leaders working together to build beneficial AI systems.
About the role
- We are looking for ML Engineers and Research Engineers to help detect and mitigate misuse of our AI systems. As a member of the Safeguards ML team, you will build systems that identify harmful use—from individual policy violations to sophisticated, coordinated attacks—and develop defenses that keep our products safe as capabilities advance. You will also work on systems that protect user wellbeing and ensure our models behave appropriately across a wide range of contexts. This work feeds directly into Anthropic's Responsible Scaling Policy commitments.
Responsibilities
- Develop classifiers to detect misuse and anomalous behavior at scale. This includes developing synthetic data pipelines for training classifiers and methods to automatically source representative evaluations to iterate on
- Build systems to monitor for harms that span multiple exchanges, such as coordinated cyber attacks and influence operations, and develop new methods for aggregating and analyzing signals across contexts
- Evaluate and improve the safety of agentic products—developing both threat models and environments to test for agentic risks, and developing and deploying mitigations for prompt injection attacks
- Conduct research on automated red-teaming, adversarial robustness, and other research that helps test for or find misuse
You may be a good fit if you
- Have 4+ years of experience in ML engineering, research engineering, or applied research, in academia or industry
- Have proficiency in Python and experience building ML systems
- Are comfortable working across the research-to-deployment pipeline, from exploratory experiments to production systems
- Are worried about misuse risks of AI systems, and want to work to mitigate them
- Have strong communication skills and ability to explain complex technical concepts to non-technical stakeholders
Strong candidates may also have experience with
- Language modeling and transformers
- Building classifiers, anomaly detection systems, or behavioral ML
- Adversarial machine learning or red-teaming
- Interpretability or probes
- Reinforcement learning
- High-performance, large-scale ML systems
- The annual compensation range for this role is listed below.
- For sales roles, the range provided is the role’s On Target Earnings ("OTE") range, meaning that the range includes both the sales commissions/sales bonuses target and annual base salary for the role.
How we're different
- We believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts. And we value impact — advancing our long-term goals of steerable, trustworthy AI — rather than work on smaller and more specific puzzles. We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science. We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time. As such, we greatly value communication skills.
- The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences.
