The Editorial Board of NATURE says ‘governments have been slow to to prioritize the ethics of AI.’ The US and China have been overly focused on the ‘the top prize,’ they say, a portion of the US$15 trillion yield expected from AI by 2030 leaving them insufficiently occupied with intergovernmental cooperation. The result puts more pressure on unilateral initiatives.
With this context they applaud a new initiative by Canada and France to advise other governments. The two counties have model proposals for the fundamentals, such as data privacy, public trust and human rights. NATURE urges caution in how they go about their work.
The new body needs to ‘follow the evidence,’ says NATURE, shielded from outside influences while respecting input from industry. More specifically, it says the Canada-France plan:
- needs the support of more countries
- must commit to openness and transparency
- must publish scientific evidence in full
- should be composed of panel members chosen for their expertise, not for which organization they represent
- should make its meetings open to observers and the media
SEE RELATED DOCUMENT
Declaration of the International Panel on Artificial Intelligence – Government of Canada, draft statement