Research Scientist CAST Propensity

the AI Security Institute; Greater London; Hybrid; GBP 65,000 - 145,000; Full time

Job summary

A leading AI research organization in London is seeking a Research Scientist to study unintended model behavior and its risks. You will engage in designing and reviewing research, applying statistical methods to data, and collaborating with experts on AI governance. Ideal candidates will have over 3 years of experience in quantitative research, strong Python skills, and the ability to communicate complex ideas effectively. This role offers a competitive salary and a chance to influence AI development at a high level.

Benefits

Learning and development stipends

Flexible working arrangements

Generous parental leave

Pension contributions

Health and wellness benefits

Qualifications

• Proven ability to operationalize key uncertainties in research.
• Knowledge of statistical methods for risk conclusions.
• Experience with production quality Python code.

Responsibilities

• Engage in research discussions and design reviews.
• Contribute to answering research science questions.
• Apply statistical methods to experimental data.

Skills

Identifying uncertainties

Statistical inference methods

Python programming

Understanding transformer architecture

Experimental design and analysis

Communication skills

Education

3 years in a quantitative research discipline

Job description

About the AI Security Institute

The AI Security Institute is the world’s largest and best‑funded team dedicated to understanding advanced AI risks and translating that knowledge into action. We’re in the heart of the UK government with direct lines to the Prime Minister’s office and work with frontier developers and governments globally.

We’re here because governments are critical for advanced AI going well, and AISI is uniquely positioned to mobilise them. With our resources, unique agility and international influence, this is the best place to shape both AI development and government action.

Applications for this position close 30^th November 2025 at 23:59.

Role Summary

Within the Cyber & Autonomous Systems Team (CAST) at AISI, the Propensity project studies unprompted or unintended model behaviour, particularly potentially dangerous behaviour: the propensity of a model to cause harm. Our current project studies the effect sizes of environmental factors on these propensities, e.g., whether models are consistently more willing to take harmful actions when their existence is threatened. We build on previous work by scaling to a range of different scenarios and variations and looking particularly for effects that are consistent throughout.

Understanding model propensities is the key missing pillar in our overall picture of risk from autonomous AI. We need empirical evidence with strong scientific credibility to justify a response to this unprecedented threat.

Responsibilities

As the second research scientist, you will:

>

Engage in discussion, write plans and designs, and review or implement code that implements those designs.

Contribute to answering research science questions such as defining when changes in scenarios are equivalent, iterating on scenario specifications, and understanding transferability to future models.

Apply statistical inference methods to draw risk‑relevant conclusions from experimental evidence.

Critically evaluate existing or proposed research methodology and adapt proposals accordingly.

Develop and iterate on Inspect tasks in Python.

Qualifications

• A proven ability to identify and operationalise key uncertainties in a research area and propose and improve experimental approaches.
• Knowledge of selecting and applying statistical inference methods for risk‑relevant conclusions.
• Experience with Python code that meets quality standards in production or collaborative settings.
• Understanding of transformer architecture and training dynamics for interpreting observable behaviour; hands‑on experience with fine‑tuning or RL is not required.
• 3 years of experience in a quantitative research discipline involving experimental design and analysis.
• Professional or educational contact with large language models and transformer theory.
• Strong written and oral communication skills.

What We Offer

Impact you couldn’t have elsewhere

• Talented, mission‑driven, supportive colleagues.
• Direct influence on how frontier AI is governed and deployed globally.
• Work with the Prime Minister’s AI Advisor and leading AI companies.
• Opportunity to shape the first and best‑resourced public‑interest research team focused on AI security.

Resources & access

• Pre‑release access to multiple frontier models and ample compute.
• Extensive operational support so you can focus on research and ship quickly.
• Collaboration with experts across national security, policy, AI research, and adjacent sciences.

Growth & autonomy

• Own important problems early if you’re talented and driven.
• 5 days of learning and development per year, stipends for learning, conferences, and external collaborations.
• Freedom to pursue research bets without product pressure.
• Opportunities to publish and collaborate externally.

Life & family

• Modern central London office (cafés, food court, gym).
• Hybrid working flexibility with occasional remote work abroad and stipends for work‑from‑home equipment.
• At least 25 days annual leave, 8 public holidays, extra team‑wide breaks, and 3 days off for volunteering.
• Generous paid parental leave (36 weeks statutory leave, shared between parents, 3 extra paid weeks, option for additional unpaid time).
• We contribute 28.97% of your base salary to your pension.
• Discounts and benefits for cycling to work, donations, and retail/gyms.
• Benefits apply to direct employees; arrangements may differ for others.

Salary

Annual salary is benchmarked to role scope and relevant experience. Offers typically land between £65,000 and £145,000, made up of a base salary plus a technical allowance. An additional 28.97% employer pension contribution is paid on the base salary.

Level range

• Level 3: Base £35,720 Technical Allowance
• Level 4: Base £42,495 Technical Allowance
• Level 5: Base £55,805 Technical Allowance
• Level 6: Base £68,770 Technical Allowance
• Level 7: Base £68,770 Technical Allowance £76,230 (total £145,000)

Selection process

The interview process may vary but typically includes:

• Initial assessment
• Initial screening call
• Research interview
• Technical assessment
• Take‑home test and interview
• Behavioural interview
• Final interview with senior leadership

In accordance with the Civil Service Commission rules, the following list contains all selection criteria for the interview process.

Key Skills

Laboratory experience, machine learning, Python, AI, bioinformatics, C/C++, R, biochemistry, research experience, natural language processing, deep learning, molecular biology.

Employment type

Full Time

Experience required: 3+ years

Vacancy count: 1

Monthly salary range: £65,000 - £145,000

EEO Statement

The Civil Service embraces diversity and promotes equal opportunities. Disability Confident Scheme and Redeployment Interview Scheme are available for eligible candidates.