PRIO and the Use of Artificial Intelligence

Artificial intelligence (AI) represents both one of the most significant opportunities and one of the most disruptive challenges facing academic research today. AI has the potential to transform how research is conducted by enabling scholars to address previously intractable questions, automate labor-intensive tasks, strengthen reproducibility and research integrity, and broaden access to scholarly outputs across linguistic and disciplinary boundaries. These developments are already happening. At the same time, it risks creating new vulnerabilities related to security, privacy, and data governance, while also eroding core scholarly competencies and intensifying pressures on publishing and funding systems. The technology’s development also carries external costs through a heavy energy and water footprint, and the concentration of capability and power in a few large technology companies. AI also raises broader questions about trust and the social foundations of scholarship. As AI increasingly contributes to the production of articles, reviews, and summaries, there is a risk that academic communication becomes detached from human judgement and accountability, potentially undermining confidence in the integrity and communal character of research itself. For PRIO, these concerns apply not only to the research we produce, but also to the two flagship journals we host. When using AI technology, PRIO expects employees to follow our general principles for due diligence, research excellence, and data security.

Our position

This sets out PRIO's position on the use of artificial intelligence and provides guidelines for employees on the responsible use of AI tools. PRIO encourages employees to make active use of AI where it can improve the efficiency and quality of our work, within the principles set out below. These guidelines do not remove the need for employees to exercise good judgement. AI technology is developing fast, and while these guidelines will be updated as AI tools evolve, everyone has a responsibility to understand the tools they use.

PRIO has three principles on the use of AI

1. The employee is accountable, not the model. AI assists, but it is not the author. Whatever we produce, publish, submit, review, or send carries our name, and we are responsible for every right, wrong or missing claim, citation, number, and line of code. PRIO expects that all AI outputs be checked.
2. Integrity is non-negotiable. Our credibility rests on honest, verifiable scholarship. AI does not change the standards for authorship, originality, or citation. Where a journal, funder, or co-author expects disclosure of AI use, we will disclose it.
3. Protect data, confidentiality, and intellectual property. Much of what passes through PRIO — personal data from research participants, unpublished manuscripts, sensitive correspondence — must not be exposed to systems we do not control. PRIO’s general information security management applies to AI unchanged.

Additional reading

• Norwegian National Committee for Research Ethics in Science and Technology (NENT) Statement on Research Ethics in Artificial Intelligence
• The European Commission’s Living Guidelines on the Responsible Use of Generative AI in Research