Artificial intelligence is in all quarters: the rise of “thinking” machines has been one of the most significant developments in the last two decades – and will only become more prominent as computing power increases.
The European Union has been working on a framework to regulate AI for quite some time, beginning in March 2018 as part of its larger Digital Decade regulations.
Work on AI regulations has been relatively slow while the EU focuses on the Digital Markets Act and the Digital Services Act, both of which aim to rein in American tech behemoths, but it is undoubtedly ongoing.
EU AI Tools
Any worthwhile legislative process should be open to criticism and analysis, and the UK-based Ada Lovelace Institute, an independent research institution focused on data policy, is giving the EU’s AI Act a thorough examination.
The full report (via TechCrunch) goes into great detail about the pros and cons of the regulation, which is a world-first, with the main takeaway being that the EU is preparing to have some pretty powerful tools at its disposal.
The EU intends to establish and empower oversight bodies that, in theory, can order the withdrawal of a high-risk AI system before requiring the model to be retrained.
The draft AI Act has been scrutinized and criticized, and it will likely fall short of the EU’s most ambitious goal: creating the conditions for “trustworthy” and “human-centric” AI.
The next battleground
The Ada Lovelace Institute report is, first and foremost, a criticism of the AI Act, since the proposal is still in the draft stage when changes can be made.
But that doesn’t mean the AI Act was a failure. In short, the report says, the AI Act is an excellent starting point for a holistic approach to AI regulation.” “However, there is no reason why the rest of the world should blindly follow an ambitious if flawed, regime that is held in place by the twin constraints of the New Legislative Framework and the legislative basis of EU law.
As computational power grows, AI will become even more integrated into people’s lives around the world than it is now.
Many of the algorithms and code powering AI are black boxes, with little to no alternatives for oversight, even when things go wrong. In the United States, for instance, efforts to regulate AI that automatically reduces benefits or dispatches police to Black and Hispanic neighborhoods have been futile.
As the world’s most powerful tech-skeptical governmental body, the EU makes sense to be leading the charge on AI regulation – and vehement criticism will only improve the result.
