Login | December 07, 2022

First they came to regulate privacy. Now the EU is after AI

Technology for Lawyers

Published: September 16, 2022

It is difficult to accurately gauge the real, on-the-ground effect of the EU privacy regulatory system GDPR. But it certainly had some effect, and has influenced privacy regulations and thought around the globe, and particularly in the US, since it went into effect in April of 2016.
And now the EU is tackling the really scary eventuality of artificial intelligence overwhelming the landscape in the not-distant future.
Could be that they watched I, Robot too many times. Or it could be that AI really is a threat.
Either way, the EU is seriously looking at reigning in AI in ways that the US is not—the EU with government regs and the US with corporate committees.
The short of it: Various forms of artificial intelligence algorithms are used as predictors of human behavior—in law enforcement, human resources and other applications.
As time has gone on, it has become obvious to close observers that these algorithms display the same prejudices as their developers—whatever that may be. And as these algos “learn,” those prejudices become stronger.
This is a problem.
In the US, the companies that develop the algos have committees studying the problem.
The rest of the world is responding with regs.
Like with data privacy regulations, the EU (and Canada) may be leading the way for the US, according to an analysis of the topic recently published by Law.com’s Legaltech news.
A little over a year ago, the EU proposed regulations to minimize the risk this technology may pose with law enforcement, safety and corporate hiring with the Artificial Intelligence Act (AIA).
I’m not going into detail, but you can look at it here: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52021PC0206
A few months ago, Canada responded with their own AI act, the Artificial Intelligence and Data Act (AIDA).
Both of those are still being worked on.
In the US, now, some legislation is being proposed.
This includes the Algorithmic Accountability Act of 2022, which would “direct the Federal Trade Commission to require impact assessments of automated decision systems and augmented critical decision processes.”
I would not bet in favor of this passing both houses or Congress by the end of this year.
So, like with the development of privacy acts around the world and the country, we will be following these developments as they arise.
You can read the full analysis here: https://www.law.com/legaltechnews/2022/07/11/as-the-u-s-grapples-with-regulating-ai-eus-approach-looms-overhead/