Eddie Ungless

Eddie Ungless

Looking for post-docs/ research scientist positions. A collection of my thoughts on the topics of ethics, fairness and social bias in NLP and beyond.

Current CV

This browser does not support PDFs. Please download the PDF to view it: Download PDF.

</embed>
CV for screenreaders Eddie Ungless [he/they] Research Interests I am currently a postdoctoral fellow at the University of Surrey researching digital harms against women MPs, including through use of generative AI. I recently received my PhD with integrated study at the University of Edinburgh, supervised by Björn Ross, Vaishak Belle and Zachary Horne. In my thesis I address how to measure bias in natural language processing (NLP) technologies using a “human-centric” approach, emphasizing human design decisions and the public’s felt experiences of these models. My research has spanned from how the public responds to biased AI technologies, to how to measure bias in a psychologically grounded fashion, to analysing the ethical risks of generative models. My linguistics and psychology background affords me a unique perspective on the topics of fairness and ethics in AI, in that I use social science theory and rigorous experimental design to gain a nuanced understanding of AI harms. I have experience conducting qualitative and quantitative research to assess the impact of AI technologies, as well as documenting the ethical risks and developing solutions, particularly to marginalised communities.

Recent Publications

† joint first author Eddie L. Ungless and Nishanth Sastry, 2026. I’m Thinking of Ending Things: Use of LLMs for Support During Break-ups. Extended Abstract at CHI, 2026. CORE: A* Eddie L. Ungless, Nina Markl and Björn Ross. 2025. Le$bean or lesbian? A survey of marginalised users' motivations for obfuscation on TikTok. Behaviour and Information Technology. Alice Ross, Cliodhna Hughes, Eddie L. Ungless and Catherine Lai. 2025. Vocal cues to perceived gender: insights from trans men Interspeech, 2025. Core: A Eddie L. Ungless, Sunipa Dev, Cynthia L. Bennett, Rebecca Gulotta, Jasmijn Bastings and Remi Denton. 2025. Amplifying Trans and Nonbinary Voices: A Community-Centred Harm Taxonomy for LLMs In Proceedings of the Association for Computational Linguistics: ACL, 2025. CORE: A* Eddie L. Ungless, Nina Markl and Björn Ross. 2025. Experiences of Censorship on TikTok Across Marginalised Identities. ICWSM, 2025. Core: A Eddie L. Ungless, Zachary Horne and Björn Ross. 2025 “Till I can get my satisfaction”: Open Questions in the Public Desire to Punish AI. Extended Abstract at CHI, 2025. CORE: A* Eddie L. Ungless, Nikolas Vitsakis, Zeerak Talat, James Garforth, Björn Ross, Arno Onken, Atoosa Kasirzadeh and Alexandra Birch. 2025. The Only Way is Ethics: A Guide to Ethical Research with Large Language Models. COLING, 2025. Atli T. Sigurgeirsson and Eddie L. Ungless†. 2024. Just Because We Camp, Doesn’t Mean We Should: The Ethics of Modelling Queer Voices. Interspeech, 2024. Core: A Charlotte Bird and Eddie L. Ungless†, and Atoosa Kasirzadeh. 2023. Ethical and Social Risks of Generative Text-to-Image Models. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society. Seraphina Goldfarb-Tarrant and Eddie L. Ungless†, Esma Balkir and Su Lin Blodgett. 2023. This prompt is measuring : evaluating bias evaluation in language models. In Findings of the Association for Computational Linguistics: ACL, 2023. CORE: A* Eddie L. Ungless, Anne Lauscher and Björn Ross. 2023. Stereotypes and Smut: The (Mis)representation of Non-cisgender Identities by Text-to-Image Models. In Findings of the Association for Computational Linguistics: ACL, 2023. CORE: A* Eddie L. Ungless, Björn Ross, and Vaishak Belle. 2023. Potential Pitfalls With Automatic Sentiment Analysis: The Example of Queerphobic Bias. Social Science Computer Review. Impact Score: 4.1

Select Invited Talks

September 2026, Lavender Linguistics, Edinburgh: Invited to be a keynote speaker at LavLang ‘26. April 2025. Loch Lomond Trust Futures Group: Invited to speak on ethical risks of adopting AI technologies March 2024. Scottish AI Summit, Edinburgh: Invited to be part of a panel on queer data and generative AI.

Research & Funding Experience

November 2025 - May 2026. Research Fellow, University of Surrey: with Nishanth Sastry. Co-authored successful UKRI ESRC funding application. Investigating gendered online abuse of women MPs. Project entails data collection, harms mapping and mitigation tool development. August - September 2025. Research Assistant, University of Surrey: with Nishanth Sastry. Produced a whitepaper on use of LLMs to seek support around major life transitions e.g. break-ups, coming out. This entailed a literature review and qualitative analysis of data from existing datasets of human-LLM interactions. October 2024 - February 2025. Student Researcher, Google Research: with Remi Denton and Sunipa Dev. Created granular taxonomy of harms to trans people from language models, based on a literature review, community survey and focus groups. Proposed pragmatic evaluation heuristics. February - July 2024; March - May 2025. Research Assistant, University of Edinburgh, Informatics School: with Alexandra Birch and Atoosa Kasirzadeh. Produced a whitepaper on how to conduct ethical research with Large Language Models, including a literature review of useful resources. Later developed into public wiki. Supported funding application for role. January - March 2023. Research Assistant, University of Edinburgh, PPLS School: with Atoosa Kasirzadeh. Helped prepare a policy document for the Department for Culture, Media & Sport on the topic of ethical risks of text-to-image models. Conducted a literature review on the risks of discrimination, exclusion and harm. 2021-23. Developer, Biascan: Developed a prototype for a system which detects implicit bias in jobs listings, and suggests alternative wording.

Teaching Experience

July 2025. Postgraduate course, University of Hamburg. Guest Lecturer on Ethics for ‘Ethics and Modern AI’. January 2023, 2024. Postgraduate course, School of Informatics, University of Edinburgh. Guest Lecturer on Ethics for ‘Doing Research in NLP’. January - April 2023, 2024. Postgraduate course, School of Informatics, University of Edinburgh. TA for ‘Case Studies in AI Ethics’ . October - December 2022, 2023. Undergraduate course, School of Informatics, University of Edinburgh. Tutor & Marker for ‘Professional Issues’ (Ethics Course). October - December 2021, 2022, 2023. Postgraduate course, School of Informatics, University of Edinburgh. TA, Tutor & Guest Lecturer for ‘Accelerated NLP’.

Professional Experience

February - August 2019. HYD, digital production agency (London, England). Deputy Head of Strategy and Innovation (Full-time). Managerial responsibilities: Responsible for day-to-day management of team of 4, including training new staff members, in addition to responsibilities of strategist role detailed below. Pitching products and services: Led pitches from conceptualisation to delivery. August 2016 - February 2019. HYD, digital production agency (London, England). Strategist then Senior Strategist (Full-time). Developing product concepts: Created digital product concepts with a research led approach. UX consultant: Worked closely with the design team to establish optimum UX.

Other Relevant Experience

• Helped organised workshop on synthetic data representation at Aarhus 2025 • Reviewer for ACL, FAccT, AIES, WiNLP (ACL), TOCHI, ICWSM • Given talks on public engagement and public speaking to CDT peers

Technical Skills

Extensive experience with Python including finetuning, and intrinsic and extrinsic evaluation of LLMs, confident using key NLP libraries such as pytorch, NLTK etc Static site design with Jekyll and Hugo; limited HTML/CCS Statistical analysis in SPSS, Smart-PLS, R Thematic coding in NVivo, Atlas

Education

2020 - 2025. PhD with integrated study in the School of Informatics. University of Edinburgh, UKRI CDT in NLP Edinburgh, Scotland. Courses: Included applied machine learning, NLP, research methods, AI ethics. 2019 - 2020. MSc - Psychological Sciences (Distinction). UCL, London, England. Courses: Included introduction to Python, experimental design, statistical analysis for psychology, qualitative analysis techniques, social psychology. Dissertation addressed default interpretation of teleology. 2013 - 2016. BA Hons - Linguistics (First Class). University of Cambridge, Cambridge, England. Courses: Included computational linguistics, generative syntax, semantics, formal logic. Dissertation addressed grammatical gender in French.