The European Commission (EC) announced on 8 April that it would launch a pilot program to ensure that ethical guidelines for the development and use of artificial intelligence (AI) can be implemented in practice.
This is the second step in the Commission’s three-part approach to the question of ethical AI, following the development of seven key requirements or guidelines for creating “trustworthy” AI developed by the High-Level Expert Group.
These include: human agency and oversight; robustness and safety; privacy and data governance; transparency; diversity, non-discrimination and fairness; societal and environmental well-being; and accountability. The Commission added that any AI that can be considered trustworthy should also respect “all applicable law and regulations”.
Industry, research institutes and public authorities have been invited to test an assessment list drafted by the group to complement the guidelines. The 52-strong panel of independent experts was appointed by the Commission in June 2018, and is comprised of representatives from industry, academia and civil society.
According to the Commission, the third and final step in its plan will be to work on building an “international consensus” on human-centric AI as “technologies, data and algorithms know no borders”.
These plans are a component of the Commission’s overarching “AI strategy”, which aims to increase public and private investments to at least €20 billion annually over the next decade in order to make more data available, foster talent and “ensure trust”.
Members of the group will present their work in detail at the third “Digital Day” in Brussels on 9 April. Following the conclusion of the pilot phase in early 2020, they will review the assessment lists for the key requirements, building on the feedback they receive, after which the Commission plans to evaluate the outcome of the project so far and propose next steps.
The Commission has also pledged to launch a set of networks of AI research excellence centres; begin setting up networks of digital innovation hubs; and together with Member States and stakeholders, start discussions to develop and implement a model for data sharing and making best use of common data spaces; before autumn 2019.
“I welcome the work undertaken by our independent experts,” Vice-President for the Digital Single Market Andrus Ansip said in a statement. “The ethical dimension of AI is not a luxury feature or an add-on. It is only with trust that our society can fully benefit from technologies.”
For Ansip, ethical AI is a “win-win proposition” that could create a “competitive advantage for Europe” should it become “a leader of human-centric AI that people can trust”.
“Today, we are taking an important step towards ethical and secure AI in the EU,” Commissioner for Digital Economy and Society Mariya Gabriel added. “We now have a solid foundation based on EU values and following an extensive and constructive engagement from many stakeholders including businesses, academia and civil society.”
The Commission is looking to put these requirements into practice while simultaneously fostering “an international discussion on human-centric AI,” she said.
AI refers to digital systems that show intelligent, human-like behaviour. By analysing their environment they can perform various tasks with some degree of autonomy to achieve specific goals, learning from data to make predictions and deliver useful insights.
The Commission estimates that the economic impact of the automation of knowledge work, robots and autonomous vehicles on the EU will reach between €6.5 and €12 trillion annually by 2025. The body has already invested what it describes as “significant amounts” in the development of AI, cognitive systems, robotics, big data, and future and emerging technologies in a bid to make Europe more competitive in this area.
This includes around €2.6 billion on AI-related areas and €700 million on research programs studying smart robots. The Commission intends to invest further in research and innovation up to and after 2020, including €20 billion per year in combined public and private investment.
However, Europe is currently behind in private investments in AI having spent €2.4 to €3.2 billion on development in 2016, compared with the €6.5 to €9.7 billion spent in Asia and €12.1 to €18.6 billion in North America.
In a press release, the Commission acknowledged that while AI has the potential to benefit a wide range of sectors – such as healthcare, climate change, law enforcement and security, and financial risk management, among others – it brings new challenges for the future of work, and raises significant legal and ethical questions.