AI Safety and Security Engineer (multiple positions)

Drive the future of AI safety and security in industrial environments across multiple specialisations, ensuring robust defenses, advanced threat modelling, and responsible AI practices.

• Strong security fundamentals: understanding threat modelling, attack surfaces, and defence-in-depth.

• Hands-on AI and machine-learning experience: able to read modern AI code and understand system internals.

• Proven engineering capability: building tools, frameworks, and automation beyond just policies.

• Experience applying a security mindset across the entire AI lifecycle.

• Ability to assess and quantify risk from agentic workflows and probabilistic AI behaviour.

• Interest in the evolving AI regulatory landscape, especially the EU AI Act.

• Collaborative approach: working effectively with AI developers and cross-functional teams.

• Fluency in English.

• Legal authorization to work and reside in Norway.

• Mentorship experience and a drive for team development.

• Ability to communicate complex probabilistic concepts to non-technical stakeholders, particularly regarding AI risk.

• Recognised contributions in AI safety or security (e.g., publications, open-source projects).

• Experience in setting evaluation standards across multi-team organisations.

• Background in industrial, energy, or similarly consequential domains.

• Proficiency in Norwegian or other Scandinavian languages.

We are scaling agentic AI, and safety & security must be first-class concerns, not afterthoughts.

These senior engineering roles are critical in defining and implementing robust offensive and defensive strategies, securing AI infrastructure, and ensuring responsible innovation in our high-stakes industrial environment.

You will apply a security mindset across the AI lifecycle, build tools and frameworks to enhance safety, and embed guardrails by design in partnership with AI developers and architects. A key aspect is assessing and quantifying risk from agentic workflows, translating probabilistic AI behavior into bounded, auditable operational risk.

Aker BP is building the foundation for agentic AI at scale. We are doing this as one operating model; strategy, architecture, science, data, safety, and delivery designed to work together, not as separate teams. This role is part of that build.

We are moving AI into the core of industrial operations, work that spans physical assets, regulated environments, and decisions that affect production, safety, and the broader energy system. We work cross-disciplinarily, with a focus on shipping into production rather than running experiments indefinitely.

We are hiring across three related tracks: offensive testing (red team), defensive engineering (responsible AI), and infrastructure security for AI systems. All three roles share a common foundation: a security mindset applied to AI systems, an understanding of modern AI threat models, and the ability to work closely with AI developers. We assess candidates against a shared senior-engineer profile and match specialisation based on strengths. Candidates are welcome to indicate a preferred track or remain open to placement.

Offensive role. Actively attempt to break model and agent behaviour across all modalities. For agents: prompt injection and reasoning-chain exploitation. For vision and acoustic models: adversarial input generation, sensor simulation, and environmental stress testing. Per-use-case risk assessment.

• Experience with penetration- and security testing of AI models and agentic workflows.

• Demonstrated adversarial machine-learning testing, prompt-injection research, or offensive security applied to AI.

• Experience with AI supply chain security (MCP and agent to agent).

Defensive and enablement role. Build responsible AI frameworks, evaluations for safety properties, and the controls that make AI systems trustworthy. EU AI Act compliance readiness and per-use-case risk frameworks are core scope. Calibration assessment and uncertainty quantification are treated as first-class safety properties alongside alignment and fairness.

• Experience with responsible AI frameworks, alignment research, or evaluation methodology for safety properties.

• Familiarity with EU AI Act compliance requirements and their engineering implications.

• Understanding of calibration assessment and uncertainty quantification for AI systems.

• Proven track record in designing and implementing safety controls for high-stakes AI applications.

• Ability to quantify and manage model uncertainty in probabilistic AI systems.

Secure our AI infrastructure, agents, and workflows. Support AI developers with guardrails and security patterns. This role sits at the intersection of cybersecurity, AI engineering, and developer enablement.

• Expertise in cloud security, infrastructure-as-code security, guardrail design, and developer-platform security.

• Experience securing agentic workflows and AI-specific attack surfaces.

• Deep understanding of secure software development lifecycle (SSDLC) for AI systems.

• Ability to embed security best practices and automated guardrails within developer workflows.

• Professional development
  We are committed to providing you with varied experience and enhancing your expertise, helping you build a strong CV for the future.

• Competitive compensation
  We offer a competitive salary, an opportunity to earn bonuses, a share savings programme, and a solid pension scheme.

• Work-life balance
  We value good health and work-life balance, offering good cafeterias, work-out facilities, and flexibility in working hours and location.

• Employee well-being
  We ensure you get support from our company health service and insurance programmes should anything happen.

• Diverse & inclusive environment
  We are convinced diversity is important for success, and everyone should feel involved, included, and respected.

• Accessible workplace
  Our offices are designed to be universally accessible for all.

• SAFER values
  Our foundation is built on being Seekers, Accountable, Foreseeable, Engaged, and Respectful in everything we do.

• AI-first mindset
  We are an AI-first company, committed to leading our industry in AI adoption and unlocking its transformative power.

• ONETEAM culture
  Cooperation is at our core; working with us means being part of our collaborative 'ONETEAM' culture.

• Safety priority
  We put safety first in all operations and expect the same commitment from every team member.

Locations: Stavanger, Lysaker

Background checks will be performed, following consent from relevant candidates. We ask that you enclose your diploma/grade transcript with your application. Furthermore, an assessment of affiliation to high-risk countries is carried out as part of the recruitment process for all positions in Aker BP.