Contact us
IT | July 22, 2022 | Write by HumanIT

New AI regulation norms in EU

Its been almost 3 years since the well-known GDPR was adopted for the EU-based companies and a lot of the tech-players became quite good with it. But as the conjuncture of the tech world becomes more and more difficult, a new more specialized norm and derivatives are being proposed by countries and different commissions to ensure that the people’s rights are still in the safe zone even after applying a new technology into the market.

A great interest in AI in the last few years provoked the appearance of the thousand of different companies that build their business strategy on the basis of artificial components and their number raises with the light-speed from year to year. But the startup world is a very rough one, so to be on the top you need to advance your algorithms quickly and with minimum expenses, which leads to a huge number of biases in the solutions that are delivered on the market. And while yet these biases can harm humanity mainly in the moral space (such as race biases, gender biases, and a kind of the very bad chatbots that know a few bad words), applying AI to the hardware technology can result in a physical threat to the human-users. Assuming this, it’s not a surprise that the EU wants to regulate this field as tight as possible to ensure the ethical and legal safety of the AI development as well as to get the dominant place in the AI race.

In just a few weeks one of the main steps will be released by EU AI Commission in the form of the first review of the Coordinated Plan on AI (insert link). This document can be the guideline that will form the AI landscape for the companies in the near future, and to be prepared for this we have gathered a few insights about the future regulation spots for you to have time to adapt your solution to them.

So, the positive side of this Plan is that it really focused on the technological development of AI solutions in the companies, that why a lot of the supportive takeaways will be implemented in this Plan, such as:

  • building a kind of the connected network of AI research centres in the EU (your company will have more possibilities to get the RnD help on the first stages of the idea validation);
  • support of the sources that will get the access to the relevant AI resources to ever EU citizens (this you will probably have more access to the high-quality data to build your solution);
  • AI solutions in the key areas will be directly supported (social startups – you will have a lot more possibilities to get your financial kick-off soon);
  • a big big sum of investment in research and education (seems like this Plan is a Xmas present for all startupers);

But what about regulations? There is no clear view about the next regulation basis in this field, however, you can take for sure that “Ethics Guidelines for Trustworthy AI” as well as the Four Deliverables presented by 52 EU experts not so long ago will strongly affect the new set of regulatory rules in AI field. To do your homework in-time you should begin to think about the next issues:

  • Is my solution enough human-centric?
  • How about sustainability, growth and fair competitiveness?
  • Are all of my developers good informed about the ethical aspects of AI? Should I bring some classes or resources for them?
  • If I operate in the Healthcare, Public Sector, and IoT- am I sure that I handle all the regulatory requirements now? And what is my vision of the AI check-up by the regulatory committee?

And one more – a new growing trend in the EU market is the kind of the business that help companies to check-up if their datasets are free of biases. Right now it looks like more prestigie label for good-fit companies with millions in revenue, but it can happened that within a few years you wont be able to register your AI solution without this kind of AI audit. Our advise- start think critically about your data right now and be responsive while gathering it, that can save you a lot of money in the future.

More precise information about the stage of the preparation and adoption of the main AI-related initiatives in the EU you can check by the link https://www.europarl.europa.eu/legislative-train/theme-a-europe-fit-for-the-digital-age/file-ai-ethical-framework. Each of the train indicates one initiative and the closest it to the borderline, the sooner the document will be adopted. Really fun and interesting way to be prepared for the serious changes in one of the most disruptive fields of technology!

And yeap, if you want to build a really ethical and socially positive solution with the top RnD devs – let us know, we will be happy to apply our expertise for your great idea.

 

Yours, HumanIT.