Cyber security specialist CISO AI model evaluation

Work on Frontier AI Systems | Via 10x.team | Major AI Lab Partner
Location Requirement: Europe (EU + UK)

10x.team is assembling a select group of European-based cybersecurity professionals and security leaders to support leading frontier AI labs in training next-generation models. This is not penetration testing. This is not SOC work. This is model shaping. You will help advanced AI systems reason better about cybersecurity: from threat modelling and incident response to security architecture, risk assessment, compliance frameworks, and the strategic complexity of defending modern enterprises.

If you value adversarial thinking, defensive precision, and the future of intelligent security systems, this is your arena.

You will contribute to Reinforcement Learning From Human Feedback (RLHF) datasets that strengthen AI reasoning across cybersecurity and information security contexts. Specifically, you will:

• Draft complex scenarios involving threat analysis, vulnerability assessment, incident response workflows, security architecture decisions, and risk management trade-offs
• Evaluate and refine AI-generated responses for technical accuracy, operational realism, appropriate severity assessment, and alignment with industry frameworks (NIST CSF, ISO 27001, MITRE ATT&CK, CIS Controls)
• Create prompt variants tailored to different contexts (e.g., CISO board briefing, SOC analyst triage, red team planning, cloud security review, third‑party risk assessment, post‑breach forensics)
• Design examples that intentionally expose weaknesses in security reasoning; missed attack vectors, flawed risk prioritisation, incomplete incident containment, over‑reliance on compliance as security
• Ensure outputs reflect European regulatory and operational realities, including NIS2, DORA, GDPR breach notification, ENISA guidance, national CERT coordination, and sector‑specific requirements (financial services, critical infrastructure, healthcare)
• Collaborate asynchronously with our research coordination team – structured, high‑focus intellectual work, distilled into training signal

• 5+ years of professional experience in cybersecurity, information security, or a CISO/security leadership role
• Based in an EU or UK jurisdiction with working knowledge of the European threat landscape and regulatory environment
• Strong expertise in at least two domains (e.g., network security, application security, cloud security, identity and access management, incident response, GRC, OT/ICS security, threat intelligence)
• Familiarity with major frameworks and standards – NIST CSF, ISO 27001/27002, MITRE ATT&CK, CIS Controls, OWASP, COBIT
• Ability to articulate security reasoning clearly across both technical and executive audiences, and identify flawed assumptions or blind spots quickly
• Current or former CISO, Head of Security, or equivalent leadership role (bonus)
• Relevant certifications (CISSP, CISM, OSCP, GIAC, CISA, CEH, or equivalent)
• Experience with European regulatory compliance – NIS2 implementation, DORA readiness, GDPR security requirements, eIDAS trust services
• Background in critical infrastructure protection (energy, finance, healthcare, transport)
• Experience in security architecture for cloud‑native or hybrid environments (AWS, Azure, GCP)
• Incident response or digital forensics experience (DFIR)
• Red team, purple team, or adversary simulation background
• Experience briefing boards, regulators, or supervisory authorities on cyber risk
• Teaching, mentoring, or thought leadership in cybersecurity

• Fully remote
• Asynchronous work (flexible scheduling – work around your operational responsibilities or personal time)
• 5–15 hours per week
• Estimated duration: 2–4+ months (extension possible)
• Independent contractor agreement – high‑quality contributors may be invited to longer‑term AI training initiatives

• Competitive hourly rate (market‑aligned for senior European cybersecurity professionals)
• Performance and quality‑driven continuation

You won't be defending one perimeter at a time. You'll be shaping how AI systems understand cybersecurity – the adversarial mindset, the defensive…