Skip to main content
Go to the home page of the European Commission (opens in new window)
English en
CORDIS - EU research results
CORDIS

Knowledge aWareness for emerging capabilities and risks in GPAI

Objective

As GPAI models grow in scale and complexity, their capabilities, i.e. the range of tasks or functions that they can perform competently, manifest as behaviours. Emerging capabilities are as yet poorly understood and characterized, leading to misconceptions about their definition, nature, predictability, and implications. Acknowledging that capability emergence is sociotechnical, not solely mathematical and computational, KnoWare aspires to make the “emerging capabilities prefiguration” the prime methodology for forecasting, explaining, assessing, and validating (pre-)emerging capabilities in GPAI systems. Prefiguration is the detectable “proto-form” of a capability as evidence that the building blocks, partial behaviours, and enabling conditions are already present, so that the full capability is plausible through scaffolds (with modest scaling, tooling, or integration). Our goal is to shift GPAI capabilities understanding from reactive observation to proactive detection. KnoWare will utilize psychometrics-inspired practices and tests and explainable AI techniques to make gradual improvement visible rather than sudden, and avoid emergence illusions. For evaluating GPAI models, KnoWare will develop a novel benchmarking approach, beyond narrow task-based performance and aggregated metrics, that considers prefigured capabilities and includes open tasks that measure not just “what the model can do” but also “whether the emerging capability is aligned with human intent” (risk-aligned governance). The approach will incorporate protocols for updating the tasks as new capabilities prefigure, also monitoring how quickly the benchmark saturates. KnoWare aims to define milestones for capability development against human values (e.g. autonomy, democratic principles, ethics) and legal compliance, monitor capability-risk co-evolution, and provide guidance for systemic risk management under the AI Act for supporting safe, trustworthy GPAI deployment in high impact sectors.

Fields of science (EuroSciVoc)

Keywords

Programme(s)

Topic(s)

Funding Scheme

HORIZON-RIA - HORIZON Research and Innovation Actions

See all projects funded under this funding scheme

Call for proposal

(opens in new window) HORIZON-CL4-2025-04

See all projects funded under this call

Coordinator

SINTEF AS
Net EU contribution
€ 853 125,00
Address
STRINDVEGEN 4
7034 Trondheim
Norway
Region
Norge Trøndelag Trøndelag
Activity type
Research Organisations
Links
Total cost
€ 853 125,00

Participants (12)