“AI, copyright and personal data – a new era of regulation?”

Recently, a Hungarian civil court referred legal questions to the CJEU concerning copyright and AI under Directive 2019/790 (see yesterday’s article in Gazeta Prawna). Meanwhile, the Polish Data Protection Authority (UODO) is expected to issue a key decision on OpenAI and the processing of personal data by ChatGPT (see today on wnp.pl). A decision is due by the end of the year.We are living in an intriguing moment—a collective attempt to “tame” AI on the internet.

Some of you may still remember the “stranger tides” that the internet represented in the early 2000s. On the biggest news portals, every article had a comment section, where users could post under pseudonyms like “Jas Śmietana” (a true demiurge of internet trolls), whose legendary comments began with “BES SĘSU” (nonsense, phonetically in Polish). Back then, identifying the author of a comment that violated someone’s personal rights, made threats, or included illegal content was nearly impossible.

This “freedom of speech” on the web arguably reached its peak on websites like “koniectvn” and “koniecpzpn.” For those who don’t remember—these were grassroots platforms for publishing (mostly unflattering) opinions about the organizations named in the URLs. Thousands of posts per day, minimal design, unfiltered language, and no moderators or administrators keeping tabs on the legality of what was being posted.

I had the professional (mis)fortune of being involved in tracking down those responsible for content on one of those sites. The servers were (figuratively speaking) in Acapulco. The Polish Act on Providing Services by Electronic Means—especially Chapter 3—haunted my dreams. Law enforcement had little idea what to do with this whole “IP address” business.And today? There are barely any comment sections left (you can maybe leave an angry emoji or a heart), and social media platforms employ armies of moderators to detect harmful or illegal content.

But a new player has entered the scene—AI. While we, mere mortals, no longer treat the online world as a terra incognita where anything goes, AI presents new challenges. This time, it’s not about emotionally charged rants smearing anyone and everyone. Now, we’re facing issues like copyright and personal data. Of course, I know that AI doesn’t have legal personality (yet—knock on wood), and that it’s ultimately people behind it. Still, AI feels like a slightly open door to doing things that humans are no longer allowed to do. Interestingly, the latter of the above said old websites is still up. For those who remember the “original” I encourage you to revisit it after 20 years and see the difference. I’m setting a reminder to return to this post in 20 years and see where we are with AI. I’m not saying “where we are with AI on the internet,” because I’m not even sure the internet—at least in the form we know it today—will still exist.

P.S. Not a single letter of this post was generated with AI.