AI Security Bootcamp is now accepting applications for the 2026 London Cohort. This is a fully funded, 7-day in-person program for experienced talent. cyber security Experts working on cutting-edge AI security.
The program will run in London from 30 August to 5 September 2026, with applications closing on 15 July. AI Security Bootcamp says applications are reviewed on a rolling basis and early applications are encouraged.
The number of participants will be 16 to 20. Successful applicants will be reimbursed for tuition fees, meals during the program, materials and accommodation in London, as well as access to travel support if required.
This bootcamp is designed for professionals with at least five years of hands-on security experience across areas such as offensive security, incident response, threat intelligence, infrastructure security, and application security. No experience with AI or machine learning required.
Participants complete pre-work to cover the baseline fundamentals of machine learning before the program, followed by a week of demos, lectures, guest speakers, and red and blue team hands-on exercises.
Curriculum covers LLM and Infrastructure Security
The AI Security Bootcamp 2026 program will focus on the threat landscape surrounding cutting-edge AI systems, from current enterprise deployments to risks associated with more sophisticated models.
Day one will cover threat modeling, including current exploitation risks, application security, infrastructure security, model theft and tampering, integrity attacks, inconsistency, and governance assurance.
Later sessions will cover adversarial examples, attacks on image models, Trojans, backdoors, fine-tuning attacks, extracting model weights, watermarking, protecting training data, and processing data during inference.
The LLM Security section includes Jailbreaking, Prompt Injection, RAG Injection, Guardrails, Tokenization Vulnerabilities, and Model Context Protocol Security.
Infrastructure content includes NVIDIA Container Toolkit exploits, GPU isolation, confidential computing, sandbox design, data center infrastructure, talent security, hardware supply chain, and governance frameworks.
The selection process includes a technical evaluation
The application process has three stages. The first stage is a short CV and application review, and AI Security Bootcamp says it will look at GitHub, past projects, and signals that contain evidence of CV.
The second stage is a technical assessment that includes threat modeling, a small PyTorch exercise, and a quick Python exercise. The third stage is a 30-minute interview that includes the applicant’s background, motivation, and short technical questions.
Selection will prioritize candidates with an interest in work involving frontier AI risks, high-consequence failure modes, or advanced threat actors. Experience with deep learning frameworks such as PyTorch is preferred but not required.
AI Security Bootcamp says approved participants must attend the entire program from Sunday, August 30th to Saturday, September 5th. Participants are encouraged to arrive on Saturday, August 29th and depart on Sunday, September 6th.
The AI Security Bootcamp says past participants are from organizations such as OpenAI, Google, Meta, Apple, Microsoft, AWS, Intel, Jane Street, Darktrace, and the Center for AI Safety.
The program page also lists the names of academic research institutions such as Stanford University, Oxford University, Cambridge University, MIT, University of California Berkeley, Tel Aviv University, CERN, National University of Singapore, and the Israel Institute of Technology Technion.
