KNOWLEDGE AND AI LAB
COGITO EPISTEMOLOGY RESEARCH CENTRE

Knowledge and AI Lab

Inquiry, understanding and knowledge in the age of artificial intelligence

​

The Knowledge and AI Lab houses COGITO's work on what artificial intelligence does to knowledge, understanding and inquiry. It is built around two projects: Emma C. Gordon's UKRI Metascience Fellowship on understanding in AI-assisted research, and Mona Simion's ERC Advanced Grant AI-KNOW on the epistemology of artificial intelligence, which begins in 2027.

​

Understanding in the Age of AI: Preserving Scientific Achievement in AI-Assisted Research Production

Dates: September 2025 – August 2027

Fellow: Emma C. Gordon

Mentor: Christoph Kelp

Policy partners: Young Academy of Europe; Initiative for Science in Europe

Funding: UKRI Metascience AI Early Career Fellowship (GBP 215,429)

​

AI now produces hypotheses, analyses and whole papers at a pace no research community can match. The fellowship asks whether AI-assisted research delivers the depth of understanding that has paradigmatically defined scientific progress, or something epistemically cheaper, and what responsible publication should look like when it does not. It runs in three stages: when AI-guided inquiry fails to generate understanding; ethical norms for AI-assisted research publication; and a publication-ethics toolkit for research institutions and policymakers, delivered with the Young Academy of Europe and the Initiative for Science in Europe. Outputs include three policy briefs, a podcast series, a joint policy workshop, an academic workshop at COGITO, and the monograph Produce More, Understand Less?.

 

Outputs so far

Books

• Emma C. Gordon, Produce More, Understand Less? AI and the Understanding That Scientific Progress Requires. The fellowship monograph, in preparation (full draft 2026).

• Emma C. Gordon, The Neuroethics of AI: An Introduction. Under contract with Polity Press, expected 2028. A case-led introduction to AI that reads the mind (neural decoding, neurorights, brain–computer interfaces) and AI that alters it (cognitive offloading, AI enhancement, AI companions); the fellowship's work on AI-assisted inquiry feeds its chapters on cognition.

Articles

• J. Adam Carter, Emma C. Gordon and Jesper Kallestrup, “What is the Epistemic Backstop Problem?” Under review (2026).

• Emma C. Gordon, invited contribution to the Topoi collection Learning and Understanding in the Age of AI (eds Christopher Bartel and Célia Teixeira), forthcoming.

Talks

• 13 July 2026: “Where does Science end and Society begin?” Public evening event, Synergy for Science 2026, SEC Armadillo, Glasgow (with J. Adam Carter).

• 24 May 2026: “Lucky Understanding Revisited” (with J. Adam Carter), What Counts as Scientific Understanding in Cognitive Science?, University of Bucharest (online).

• 31 March 2026: “AI, Scientific Progress and Understanding”, AI Epistemics: Knowing in the Age of AI, Onassis Library, Athens.

• February 2026: Keynote, “The Epistemic AI Alignment Problem” (with J. Adam Carter), Engineering the Concept of Hermeneutical Injustice, University of Seville.

• 27 October 2025: “AI, Scientific Research Culture and Progress” (with J. Adam Carter), University of Glasgow Festival of Data Science and AI.

• 16 October 2025: “AI, Inquiry, and Scientific Progress” (with J. Adam Carter), Inaugural Inquiry Lecture of the Young Academy of Europe, Barcelona.

Knowledge exchange

• Young Academy of Europe: Board member and Chair of the Selection Committee; the 2025–26 YAE Inquiry Lecture series on AI opened with the Gordon and Carter lecture, and Gordon chaired its February 2026 webinar.

• Elected to the Young Academy of Scotland, 2026.

• Research cited in the United Nations Human Development Report 2025 (Gordon and Seth, PLOS Biology, 2024, on the ethics of brain–computer interfaces for cognitive enhancement).

• Policy briefs and the publication-ethics toolkit with the Young Academy of Europe and the Initiative for Science in Europe: in progress, 2026–27.

Links: Talks and public engagement · Publications

​

AI-KNOW: The Epistemology of Artificial Intelligence: A Foundational Framework

Dates: 2027–2032

Principal Investigator: Mona Simion (University of Oxford; co-founder and Honorary Director of COGITO)

Funding: European Research Council Advanced Grant (EUR 2.5 million)

​

AI-KNOW asks whether AI systems should be understood merely as sophisticated tools or as genuine epistemic agents capable of knowledge, understanding, justification and epistemic responsibility. It works across four areas: the nature of artificial epistemic agency; the conditions under which AI systems can generate knowledge and understanding; the principles governing trust and collaboration between humans and AI; and the practical implications for medical diagnosis, legal decision-making, education and disinformation detection. The project starts in April 2027 at Exeter College, Oxford. Announcement.