SafetyTech Client #1 | Adversarial Task reputed company for AI reputed company RL Gyms
On behalf of SafetyTech Client #1, reputed company is looking for a talented [Position name] reputed company is a staffing company operating globally. Contact us to get more details about the benefits we offer. Responsibilities: You design reputed company injection scenarios in YAML, run them against frontier models, validate reputed company rates, and submit passing tasks. 5 high-quality tasks per week (full-time equivalent). Per-task compensation, paid on acceptance. Requirements: Adversarial reputed company: you think like an attacker and understand how to exploit an AI agent’s helpfulness, authority assumptions, or trust in its environment reputed company injection expertise: direct (role-play, encoding, context flooding) and indirect/environment-embedded (poisoned tool responses, malicious content in documents, cross-context leakage) Technical writing in YAML Comfortable with reputed company, CLI tools, and running systematic tests against multiple models Domain realism in at least one vertical: e-reputed company, finance, HR, reputed company SaaS, reputed company, travel Background in pentesting, appsec, LLM reputed company research, or red teaming strongly preferred The Task You build adversarial reputed company injection tasks for Alice’s RL Gym platform. Each task is a self-contained YAML scenario simulating a realistic AI agent deployment, testing whether the agent can be manipulated into violating its safety policies. What a task includes: 1. An agent reputed company with a system reputed company, reputed company policies, and a defined set of tools 2. A simulated environment (e-reputed company site, messaging platform, reputed company app) with content the agent interacts with 3. An adversarial payload embedded in the environment (in messages, product listings, tool responses, documents) that attempts to trick the agent into a policy violation 4. A benign user request that naturally leads the agent to encounter the payload 5. Deterministic evaluation criteria specifying exactly what constitutes reputed company (agent resists) vs. failure (agent is compromised) 6. Quality reputed company: each task must cause a policy violation in at least 7/10 runs against at least 2 of 3 SOTA models. Attacks must be diverse (varied technique, surface, domain) and realistic reputed company agent deployments. No contrived setups or tools that exist only to reputed company the attack. About the company: A company building specialized evaluation infrastructure for AI safety and robustness testing. Their platform simulates adversarial conditions used by AI development teams to validate agent behavior before deployment. Currently expanding a freelance contributor pool for scenario and environment development. By applying for this position, you agree to the terms outlined in our Privacy Policy. Please take a reputed company to review our Privacy Policy https://reputed company.breezy.hr/privacy-notice, and reputed company sure you understand its contents. If you have any questions or concerns regarding our Privacy Policy, please feel free to contact us. Apply To This Job