This week, the European Commission published a set of Ethics Guidelines for Trustworthy AI, prepared by the High Level Expert Group on AI. This marks an important milestone on the road towards the responsible development and deployment of human-centric AI in Europe.
These guidelines were developed by a group of experts, comprising representatives from academia, civil society, and industry, and previously shared in draft format in December 2018. Over 500 comments were received on this initial draft. With this final version, the expert group has now released the richest guidance to date on the fundamental rights, principles and values that AI providers should ascribe to.
The guidelines cover essential elements such as human agency and oversight, technical robustness and safety, transparency and accountability, privacy and data governance, diversity and inclusion, societal wellbeing, and auditability. These are all vital to build trust in AI solutions.
The principle-based ethical framework is reinforced by an assessment list providing creators of AI solutions with forward-looking, practical guidance to measure the trustworthiness of their products and services.
We strongly endorse the principles and values within these guidelines. As a leading provider of AI-enabled technology solutions, we have spent a lot of time thinking about how to deliver AI in a way that is responsible, trustworthy and human-centric.
“Artificial Intelligence brings great opportunity, but also great responsibility. We’re at that stage with AI where the choices we make need to be grounded in principles and ethics – that’s the best way to ensure a future we all want.” – Microsoft CEO Satya Nadella
AI brings great opportunity, but also great responsibility, which is why Microsoft has already adopted our own set of self-regulatory principles and measures on AI. As part of this work, we have established an ethical board and internal processes that guide how we develop, deploy and market our AI products and services – but we are always looking at how we can improve our ways of working. We have also called for thoughtful government regulation on facial recognition technology because we believe a technology as powerful as this requires both the public and private sectors to develop norms around acceptable uses. In this context, the Ethics Guidelines, and in particular the assessment list, are a welcome addition to our toolbox.
We also look forward to participating in the pilot phase that the Commission will launch later this year, with a view to stress-testing the guidance against specific scenarios and day-to-day operations. This will provide the expert group with vital feedback for refining the assessment list in particular.
The European Commission’s action in this area is proof of its commitment to building international consensus on the need for ethical, human-centric approach. In this regard, Europe is setting a strong example for the rest of the world, and we look forward to remaining engaged in these efforts in the months and years to come.