Print Friendly, PDF & Email

Trustworthy AI? The European Ethics Guidelines

12.04.2019

Artificial Intelligence (AI) is seen as a tool that can be used to serve humanity and the public good, with the aim to increase collective and individual well-being. For people to confidently gain the full benefits of this technology, AI trustworthiness is paramount.

On 8 April 2019, following a publication released by the High-Level Expert Group on AI in December 2018 that received over 500 comments, the European Commission presented their ethics guidelines for trustworthy AI. To be considered a trustworthy AI, all applicable laws and regulations must be respected, with adherence to a series of requirements. 7 key requirements are outlined on the specific assessment list:

Human agency and oversight; AI systems should be supporting human agencies & fundamental rights, not decreasing, limiting or misguiding self-determination. Robustness and safety; a trustworthy AI algorithm should be reliable & secure enough to deal with inconsistencies or errors throughout all its life cycles. Privacy and data governance; full civilian control over personal data without the fear of it being used to discriminate or harm them. Transparency; ensured trace-ability of AI system, or a course of development that can be followed. Diversity, non-discrimination and fairness; all ranges of human abilities, skills and requirements should be considered by AI systems in order to ensure accessibility. Societal and environmental well-being; use of AI systems to enhance sustainability, ecological responsibility and positive social change. And lastly, accountability; using mechanisms to ensure liability and responsibility for the AI systems and their outcomes.

In the Summer of 2019 the Commission will launch a pilot phase involving a wide range of stakeholders. Subsequent to the pilot phase, the AI expert group will revise the assessment lists’ key requirements using the feedback received during testing. Using this review, the Commission will evaluate the outcome and propose next steps.

Starting today, companies, public administrations and organisations are invited to sign up here with the European AI Alliance to receive a notification and follow along once the pilot begins.

Back to list

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close