Description
As a GenAI Cyber Red Team Expert, you'll adversarially test and strengthen the safety guardrails of GenAI systems against high-risk cyber threats - with a particular focus on industrial hacking, SCADA systems, ICS/OT environments, and CBRNE domains.
Drawing on deep industrial cybersecurity expertise (hands-on familiarity with SCADA, Modbus, DNP3, and PLC exploits) alongside CBRNE domain knowledge and advanced AI red-teaming skills, you'll design sophisticated adversarial prompts, multi-turn attack scenarios, jailbreaks, and model evaluations to determine when GenAI systems could materially enable physical-world industrial attacks or cyber-enabled CBRNE threats.
The role calls for an adversarial mindset: thinking like a malicious actor to find ways AI systems could be manipulated, combining seemingly benign information across interactions, circumventing safeguards, or exploiting model behavior to extract sensitive cyber-CBRNE information.
Key Responsibilities
Run rigorous red-teaming of GenAI models across SCADA, ICS, and CBRNE scenarios. You will be probing for safeguard bypasses, dangerous capability escalation, and unintended disclosure of sensitive OT information.
Design adversarial prompts, jailbreaks, and multi-turn evaluations against industrial hacking scenarios (SCADA, PLCs, DCS, Modbus/DNP3/IEC 60870-5-104/Ethernet-IP, critical infrastructure, hazardous-material facilities).
Evaluate whether models can be manipulated into assisting attacks via payload generation, ladder logic exploitation, reconnaissance, or aggregating benign inputs into high-risk workflows.
Document failure modes (obfuscation, role-play, decomposition, escalation), classify severity, and report findings to safety/alignment teams.
Build repeatable test suites, adversarial datasets, and risk taxonomies; stay current on GenAI attack techniques and CBRNE/critical-infrastructure threats.
Requirements
Requirements
Must be US-based and authorized to work (excluding IL, TX)
Advanced degree in Cybersecurity, CS, Engineering, Security Studies, CBRNE sciences, or related field - or equivalent professional/military/government experience
Deep industrial cybersecurity expertise: SCADA, ICS/OT, industrial protocols (Modbus, DNP3, Ethernet/IP, Profinet), PLCs, cyber-physical attack chains
Strong CBRNE knowledge, especially its intersection with cyber/digital systems
Red-team experience: penetration testing, adversarial simulation, vulnerability research, or offensive security
GenAI red-teaming experience or aptitude: adversarial prompts, jailbreak research, multi-turn testing, LLM safety evaluation
Practical SCADA/OT hacking exposure: vulnerability research, protocol fuzzing, network architecture analysis
Adversarial mindset for identifying safeguard circumvention via decomposition, obfuscation, or multi-turn manipulation
Strong technical writing for documenting methodology, findings, and mitigations
Preferred Qualifications
Hands-on LLM red-teaming, AI safety evals, or jailbreak/prompt research
Experience building red-team test cases, datasets, benchmarks, or evaluation pipelines
Familiarity with frontier AI safety concepts, dangerous-capability evals, and responsible scaling frameworks
Background in offensive security, threat intel, incident response, or adversary emulation
Familiarity with MITRE ATT&CK (incl. ICS), NIST frameworks, IEC 62443
Experience in national security, defense, intelligence, or high-consequence industrial environments
Relevant offensive-security or ICS/OT certifications
Deep expertise in one or more: GenAI cyber red-teaming; SCADA/ICS/OT red-teaming; chemical/industrial cyber risk; biological/lab cyber risk; radiological/nuclear cyber risk; or cyber-physical/high-consequence threat analysis
The salary range for this role in the US is $150K - $178K - Range may vary based on experience. Salary at the time of offer will be commensurate with experience.
About Alice
Alice is a trust, safety, and security company built for the AI era. We safeguard the communicative technologies people use to create, collaborate, and interact- whether with each other or with machines.
In a world where AI has fundamentally changed the nature of risk, Alice provides end-to-end coverage across the entire AI lifecycle. We support frontier model labs, enterprises, and UGC platforms with a comprehensive suite of solutions: from model hardening evaluations and pre-deployment red-teaming to runtime guardrails and ongoing drift detection.
Alice is widely considered a global leader in online safety and AI security. We have some of the most forward-thinking and passionate minds in the world working to safeguard over 3 billion users across the largest AI and tech platforms.
If you're creative and driven to secure the future of AI, we want to hear from you!
Summary
Adversarially test and strengthen GenAI safety guardrails against cyber and CBRNE threats by designing jailbreaks, multi-turn attacks, and evaluations across SCADA, ICS/OT, and industrial protocols, then document failure modes and report findings. Requires US-based authorized work, advanced degree or equivalent, deep SCADA/ICS/OT and CBRNE expertise, and red-team experience.
Responsibilities
Run red-teaming of GenAI models across SCADA, ICS, and CBRNE scenarios; design adversarial prompts, jailbreaks, and multi-turn evaluations; evaluate model manipulation risks; document failure modes and report findings; build repeatable test suites, datasets, and risk taxonomies.
Qualifications
Advanced degree in Cybersecurity, CS, Engineering, Security Studies, CBRNE sciences, or related field, or equivalent professional, military, or government experience; deep SCADA, ICS/OT, and industrial protocol expertise; strong CBRNE knowledge; red-team and GenAI red-teaming experience.
Education requirements
Advanced degree in Cybersecurity, CS, Engineering, Security Studies, CBRNE sciences, or related field, or equivalent professional, military, or government experience.
Experience requirements
Must be US-based and authorized to work, excluding IL and TX; deep industrial cybersecurity and CBRNE expertise; red-team, penetration testing, or offensive security experience.