FUTURE-AI guideline

In this page, we provide definitions and justifications for each of the six guiding principles and give an overview of the FUTURE-AI recommendations.

The following table provides a summary of the recommendations, together with the proposed level of compliance (i.e. recommended vs. highly recommended).

^{List of the FUTURE-AI recommendations, together with the expected compliance for both proof-of-concept (Low ML-TRL) and deployable (High ML-TRL) AI tools (+: Recommended, ++: Highly recommended).}

		Recommendations	Low ML-TRL	High ML-TRL
F	1	Define any potential sources of bias from an early stage	++	++
	2	Collect data on individuals’ attributes, when possible	+	+
	3	Evaluate potential biases and bias correction measures	+	++
U	1	Define intended clinical settings and cross-setting variations	++	++
	2	Use community-defined standards (e.g. clinical definitions, technical standards)	+	+
	3	Evaluate using external datasets and/or multiple sites	++	++
	4	Evaluate and demonstrate local clinical validity	+	++
T	1	Implement a risk management process throughout the AI lifecycle	+	++
	2	Provide documentation (e.g. technical, clinical)	++	++
	3	Define mechanisms for quality control of the AI inputs and outputs	+	++
	4	Implement a system for periodic auditing and updating	+	++
	5	Implement a logging system for usage recording	+	++
	6	Establish mechanisms for human oversight and governance	+	++
U	1	Define intended use and user requirements from an early stage	++	++
	2	Provide training materials and activities (e.g. tutorials, hands-on sessions)	+	++
	3	Evaluate user experience and acceptance with independent end-users	+	++
	4	Evaluate clinical utility and safety (e.g. effectiveness, harm, cost-benefit)	+	++
R	1	Define sources of data variation from an early stage	++	++
	2	Train with representative real-world data	++	++
	3	Evaluate and optimise robustness against real-world variations	++	++
E	1	Define the need and requirements for explainability with end-users	++	++
E	2	Evaluate explainability with end-users (e.g. correctness, impact on users)	+	+
General	1	Engage inter-disciplinary stakeholders throughout the AI lifecycle	++	++
	2	Implement measures for data privacy and security	++	++
	3	Define adequate evaluation plan (e.g. datasets, metrics, reference methods)	++	++
	4	Identify and comply with applicable AI regulatory requirements	+	++
	5	Investigate and address ethical issues	+	++
	6	Investigate and address social and societal issues	+	+