Perspectives: artificial intelligence and scientific autonomy: a policy agenda for emerging research systems

Abstract Artificial intelligence is becoming part of the infrastructure through which scientific knowledge is produced, yet access to computing, data, advanced models and expertise remains highly uneven. This Perspective argues that, for emerging research systems, the central science-policy challenge is not simply AI adoption but scientific autonomy: the capacity to choose, understand, adapt, evaluate and govern the technologies on which research increasingly depends. Scientific autonomy is not technological sel…

By the Truvace deskOctober 22, 20262 min read1 sourcePeer-reviewedTRV-2026-2178· cite & verify

In brief

This peer-reviewed Perspective, published in Science and Public Policy, argues that AI has become part of the infrastructure used to produce scientific knowledge. It reframes the policy question for emerging research systems from whether to adopt AI to whether they can retain real control over the tools their research relies on.

The authors use Brazil as an illustrative case to develop a framework and a policy agenda. The matter is significant because AI could speed discovery while leaving poorer or smaller research systems further behind. The piece is a conceptual argument rather than a measured evaluation, so it does not show that the proposed measures work.

Main points

Scientific autonomy is defined as the capacity to choose, understand, adapt, evaluate and govern the technologies research depends on.
The authors say autonomy is strategic interdependence supported by domestic capabilities, not technological self-sufficiency.
Four dimensions are proposed: epistemic, infrastructural, human-capability and governance autonomy.
The six-part agenda covers shared computing, research-data infrastructure, AI literacy, interdisciplinary networks, adaptive governance and capability-building international cooperation.

The problem

Access to computing, data, advanced models and expertise is highly uneven, so emerging research systems risk depending on technologies they cannot choose, adapt or govern, deepening asymmetries between research systems.

The rundown

The Perspective starts from the observation that AI is no longer only a tool but part of the infrastructure through which scientific knowledge is produced, while access to computing, data, advanced models and expertise remains highly uneven across research systems.

From this, the authors say the central challenge for emerging research systems is scientific autonomy rather than adoption alone. They separate autonomy from self-sufficiency, describing it as strategic interdependence backed by domestic capabilities.

Using Brazil as an illustrative case, they set out four dimensions of autonomy and a six-part policy agenda. The warning is conditional: absent deliberate capacity-building, AI may accelerate discovery while deepening structural asymmetries.