Lisa Thiergart, Luis Cosio (SL5 Task Force)
The SL5 Task Force will build out a prototype SL5 datacenter this year together with frontier AI labs. This will be a massive research and engineering project with many avenues for spinning out new organizations and research programs. This project is urgent due to this technology being needed in the next 1 to 2 years.
Stream overview
We are interested in supporting research leaders capable of and interested in growing towards being high impact founders or project leads within AI Security areas relevant to SL5. Specifically:
- physical security of datacenters (includes actually going to our datacenter and physically testing the components, contributing to AI705 research & publication)
- air gap implementation & benchmarking researcher productivity
- supply chain research for hardware and software procurement for SL5 datacenter construction
- AI-specific vulnerability & threat research
- Verification & Attestation of software and hardware components
- Designing & Building new simple custom hardware components where necessary
- Developing red teaming frameworks, infrastructure and standardization at the SL5 level
- standards work in high security datacenters (next iteration of SL5 standard, publishing implementation guidance, working with NIST and lab stakeholders to express voluntary industry standards)
- working directly with frontier labs & affiliated teams to develop implementable and cost-optimized security designs & implementations
SF Bay Area
Mentors
Lisa is the CEO of Security Level 5, an AI Security Tech Lab she founded with the mission to create the technical and strategic optionality for frontier AI labs to reach SL5 (security against priority nation state attacks) for their core internal operations in the coming years. Her team developed the world’s first SL5 Standard, and is currently prototyping mock SL5 datacenters in coordination with frontier AI labs. The SL5 work brought together 100+ security engineering specialists across frontier AI labs, the US intelligence community and broader AI Security ecosystem to chart the technical path towards reaching nation-state secure AI Datacenters and frontier AI workflows by 2028. Lisa’s background includes a BSc in Computer Science from TUM, Graduate Machine Learning Research at Georgia Tech, as well as an AI Security DPhil researcher affiliation at the Oxford HAIGL Lab. Previous to founding SL5, Lisa was a senior director at IST, research lead at MIRI where she founded and led the technical governance team (verification, hardware security, etc), as well as participated in MATS3 as a mentee under Alex Turner in 2023 where she helped pioneer the technique of activation steering. Lisa has been a 2026 FLI Fellow, 2025 Brains Fellow, 2024 Foresight Fellow, Fulbright Scholar and participant in Entrepreneur First and 5050 by 50Years.
I work at the intersection of frontier AI and high-security systems. My focus is turning abstract safety and security requirements into deployable technical artifacts that can withstand real adversaries, from nation-state attacks to loss-of-control scenarios.
Focus areas:
- Frontier AI Security (SL5): Securing model weights, inference systems, and lab infrastructure against nation-state threats and loss-of-control failure modes.
- AI Alignment and Interpretability: Mechanistic interpretability, scalable oversight, and monitoring-based interventions.
- Security Engineering: Air-gapped deployments, attack surface minimization, and high-assurance runtime design.
Fellows we are looking for
This stream is best for strong technical IC's looking to move into research lead / tech lead / org lead positions in the future.
Essential:
- BSc Computer Science or Electrical engineering, or comparable experience
- 3+ years of Security Engineering, Infrastructure Engineering or related experience - demonstrated ability to independently lead complex projects as an IC relating to designing, operationalising and implementing security programs/controls.
- demonstrated previous research lead / tech lead / management experience (eg having lead at least one successful project with 2+ people before that involved shipping rapidly and on novel technical ground)
- strong python/rust skills (eg independently debugging project code, etc) OR
- excellent written and verbal communication skills for technical and non-technical collaborations
- comfort adopting highly automated workflows (claude code, etc)
Preferred:
- AI research experience (eg have previously debugged a training run, have worked on interpretability research on large models, strong conceptual understanding of most major alignment research directions)
- basic familiarity with common industry security standards and guidelines (NIST SP800-53, SOC, ICD705, etc)
- Hands on hardware and physical security experience, such as having been involved with TEMPEST / SCIF construction / Physical security of a datacenter build
- Procurement and Supply Chain Research Experience
Not a good fit:
- Those not open to potentially working in-person in the Bay area after MATS
- Those who need heavily structured guidance (I expect scholars to propose and iterate on ideas independently)
- Those not interested in potentially taking on management or leadership responsibility in the future (eg. this stream aims to train people who might have future jobs that involve 10-50% time on leadership tasks and 50-90% time on technical individual contributor time).