PhD Candidate, Computer Science, UCL, UK

Research Interests
- Intent in AI
- Control of Autonomous Agents
- Silicon Collar Crime
- Causal inference
- Law and AI
- Fraudulent Financial Statement detection
- Herding and informational cascades
- Regulatory/Research/University capture
- Failure of RL
Publications
- Ashton, H. (2020) Definitions of intent for AI derived from common law, 14th International Workshop on Juris-informatics (JURISIN 2020)
- Ashton, H (2021) Campbell-Goodhart’s law and Reinforcement Learning, ICAART 2021
- Ashton, H (2020) AI Legal Counsel to train and regulate legally constrained Autonomous systems, 4th annual workshop of applications of artificial intelligence in the legal industry, IEEE BigData 2020
- Ashton H (2021) Definitions of intent suitable for algorithms (pre-print)
- Ashton H (2021) Extending counterfactual accounts of intent to include oblique intent (pre-print)
- Ashton, H (2022) Defining and Identifying the legal culpability of side effects using causal graphs, AAAI workshop on Artificial Intelligence Safety (SafeAI22)
- Ashton, H and Franklin, M (2022) The problem of behaviour and preference manipulation in AI systems, AAAI workshop on Artificial Intelligence Safety (SafeAI22)
- Ashton, H, Franklin, M and Lagnado, D (Forthcoming) Testing a definition of intent in a legal setting (preprint)
- Franklin, M; Ashton, H; Gorman, R and Armstrong, S (2022) Missing mechanisms of manipulation in the EU AI Act, 35th International Conference of Florida Artificial Intelligence Research Society (FLAIRS)
- Franklin, M; Ashton, H; Awad, E and Lagnado, D (2022) Causal Framework of AI Responsibility, Fifth AAAI/ACM Conference on Artificial Intelligence, Ethics and Society
Teaching
- 2019/20 – Algorithmic Trading PGTA
- 2020/21 – Market Microstructure PGTA
Presentations & Writing
- Challenges in Algorithmic Crime in Finance, GovTech Lab, UCL, May 2019
- Emergence of spoofing in RL traders, Transfer Viva, UCL, Jan 2019
- Jurisin 2020 Presentation , November, 2020