The Importance of Trustworthy Artificial Intelligence
Researchers at Linköping University have developed a roadmap, intended to guide policy makers towards the future of trustworthy artificial intelligence.
Artificial intelligence (AI) is increasingly present in our daily lives, and it is believed that this is just the beginning. For this to continue, however, requires ensuring that the AI is trustworthy in all scenarios. To help in this endeavour, Linköping University (LiU) is coordinating TAILOR, a European project that has developed a research-based roadmap to guide research funding bodies and policy makers towards reliable AI in the future. “TAILOR” is short for Foundations of Trustworthy AI – Integration, Learning, Optimization and Reasoning.
About the TAILOR project
Funded by EU Horizon 2020, TAILOR is one of six research projects created to develop the AI of the future. TAILOR lays the foundation for trustworthy AI, producing a framework, guidelines and specification of the needs of the AI research community.
The roadmap presented by the project is the first step on the road to standardization, where policy makers and research funding bodies can gain an understanding of the development of trustworthy AI. However, research issues must be resolved before this can be achieved.
Fredrik Heintz, Professor of Artificial Intelligence at LiU and coordinator of the TAILOR project, highlighted the importance of trustworthy AI, explaining: “The development of artificial intelligence is in its infancy. When we think back to what we are doing today 50 years from now, we find it quite primitive. In other words, the essence of the field remains to be discovered. That’s why it’s important to lay the groundwork for reliable AI now. »
Three criteria for a trustworthy AI have been defined by the researchers: it must meet several ethical concerns, it must comply with the laws and its implementation must be robust and safe. However, these criteria pose challenges, particularly the implementation of ethical principles.
Heintz explained, “Take justice, for example. Does this mean an equal distribution of resources or that all actors receive the necessary resources to bring them all to the same level? We face major long-term questions, and it will take time to answer them. Remember that the definition of justice has been debated by philosophers and scholars for hundreds of years.
Basic research on AI is a priority
Big, comprehensive questions will be at the center of the project and standards will be developed for everyone working with AI. However, Heintz believes this can only be achieved if basic AI research is a top priority.
“People often see AI as a technology problem, but what’s really important is whether we get any societal benefits from it. If we’re going to get AI that’s trustworthy and works well in society, we need to make sure it’s people-centric,” Heintz said.
Several legislative proposals drafted within the EU and its member states are drafted by lawyers, but they are believed to lack specialist knowledge within AI – a serious problem according to Heintz.
“Legislation and standards must be knowledge-based. This is where we researchers can contribute, providing information on the current cutting edge of research and enabling informed decisions. It is important that experts have the possibility of influencing questions of this type.
The full roadmap is available at Trustworthy AI Research and Innovation Strategic Roadmap.
