Generative AI in science: European Parliament study proposes ten policy options to tackle five key risks

A study published in October 2026 by the European Parliament’s Panel for the Future of Science and Technology (STOA) set out ten policy options to address the risks raised by generative AI in scientific exchange in the EU.

Generative AI could make research more accessible, findable and usable, but it could also weaken scientific exchange by “obscuring provenance, deepening dependency on opaque infrastructures” and increasing “synthetic overload”, the study said.

The authors identified five priority risks: slow adaptation of governance, loss of trustworthiness of scientific outputs, lack of reproducibility, inequity of access and erosion of Europe’s strategic competitiveness.

To address them, they put forward pilot governance programmes and an EU observatory on scientific exchange; AI-use…