AI ethics and governance has change into a loud area.
Ultimately rely, the OECD tracker counts over 1,800 national-level paperwork on initiatives, insurance policies, frameworks, and techniques as of September, 2024 (and there appears to be consultants and influencers opining on each one).
Nevertheless, as Mittelstadt (2021) succinctly places in a manner that solely tutorial understatement can, rules alone can’t assure moral AI.
Regardless of the abundance of high-level steerage, there stays a notable hole between coverage and real-world implementation. However why is that this the case, and the way ought to knowledge science and AI leaders give it some thought?
On this collection, I goal to advance the maturity of sensible AI ethics and governance inside organizations by breaking down this hole into three elements, and draw from analysis and actual world expertise to suggest methods and constructions which have labored in implementing AI ethics and governance capabilities at scale.
The primary hole I cowl is the interpretation hole, which arises from the problem of making use of rules expressed in obscure language resembling ‘human centricity’ and…
