What would it take to trust AI?

While the importance of new technologies has been recognized as critical to achieving the SDGs, AI is expected to play a particularly important role in accelerating progress towards each one of the 17 SDGs. Lives across the world can be improved in fundamental ways.  AI has the potential to help address many of humanity’s most critical social issues, including those related to health, education, nutrition, disability, youth, social inclusion, and poverty.

The transformative power of AI also brings with it complicated chal­lenges, ranging from technical to ethical and human rights issues, to security risks, to socio-economic issues such as the disruptive impact of AI on employment.  There are concerns that yet another digital divide may be opening, one with profound implications for inequality, unless the needs of both developing and the least developed countries and all segments of the population are considered.

Central to these concerns is the issue of Trust.  

As the systems upon which our society depends become more autonomous and complex, and operate with less human supervision, the implications of risks to security and privacy, as well as transparency and accountability – in both, AI algorithms and tools, and the data they depend on - could become greater. This is especially true in mission critical applications, such as those in transportation, healthcare, communication and many other sectors, taking into account the increasing adoption of applications such as automated driving systems, intelligent drones, healthcare robots etc.  

Therefore, as the capabilities of AI progress from narrow to general, it will be critical that systems and policies are in place to ensure trust. This High Level Dialogue will discuss policy imperatives to better inform strategies to build safeguards to ensure trust in AI for maximizing the benefits, and enabling as opposed to stifling innovation through balanced and considered policies.