If last week’s posts were any indication, Artificial Intelligence has been busy testing our human intelligence. It felt like the digital age sat us all down for a surprise exam, one we took entirely through our screens, only stepping away for some guitar riffs, a bite of wine and cheese and the occasional patent case. Here’s what you might have missed while The Kats were busy proving that humans still have the upper hand (for now):
Georgia Jenkins then tuned our attention to Spotify’s latest AI-infused move. She broke down the platform’s integration into ChatGPT, a slick innovation promising “personalized discovery” that might actually tighten algorithmic control over what we hear. By linking chatbots and streaming, she showed how the line between recommendation and manipulation blurred, leaving artist pay and listener choice on shuffle. Many readers (this Kat included) were left questioning whether their Discover Weekly really knew them, or just who paid to sound like it did.
Eleonora Rosati reminded us that text and data mining (TDM) exceptions aren’t a carte blanche for AI developers. They only cover narrow acts like extraction or reproduction not the full scope of model training. By revisiting the three-step test and fair dealing rules, she showed why unlicensed AI training can’t sneak by: innovation still needs permission. With global licensing deals already shaping this space, Eleonora's message was clear; exceptions might crack the door open, but licensing still holds the keys.
Midway through their digital deep dive, the Kats stumbled on a true AI brain-twister: if certain AI systems are banned under the EU AI Act, does that mean they can’t be patented? As it turns out, not quite. Our Katfriends Andreas Engel and David Faber explained the Act’s prohibitions may sound like red lights, but with exceptions for research and development, and plenty of case-by-case nuance, innovation isn’t entirely boxed in. Regulation and patent law, it seems, can still coexist proving once again that even in a world ruled by algorithms, a little human reasoning goes a long way.
Copyright and AI
Söğüt Atilla-Aydın cranked up the volume in a case where the IPEC had to decide who really authored Venom’s demonic logos and album art. Between bandmates, blurred memories, and misspelled nicknames, the court struck a chord: originality isn't about how loud the art screamed, but whether it shows genuine creative freedom, even when it is just sharpening the edge of a letter.Georgia Jenkins then tuned our attention to Spotify’s latest AI-infused move. She broke down the platform’s integration into ChatGPT, a slick innovation promising “personalized discovery” that might actually tighten algorithmic control over what we hear. By linking chatbots and streaming, she showed how the line between recommendation and manipulation blurred, leaving artist pay and listener choice on shuffle. Many readers (this Kat included) were left questioning whether their Discover Weekly really knew them, or just who paid to sound like it did.
Eleonora Rosati reminded us that text and data mining (TDM) exceptions aren’t a carte blanche for AI developers. They only cover narrow acts like extraction or reproduction not the full scope of model training. By revisiting the three-step test and fair dealing rules, she showed why unlicensed AI training can’t sneak by: innovation still needs permission. With global licensing deals already shaping this space, Eleonora's message was clear; exceptions might crack the door open, but licensing still holds the keys.
Katfriend Lea Kristin Fischer might have cheated on the Kats’ exam, and passed with flying colours. Her post on the German Federal Court of Justice’s Action Replay II decision showed that, in the software world, breaking the rules isn’t always breaking the law. The court found that “cheat software” doesn’t infringe copyright if the game’s code stays untouched. With AG Szpunar’s analogy, skipping to the end of a detective novel may spoil the fun, but it’s not illegal. Fischer proved that even a little cheating can still be a fair play.
Patents
In his guest post Greg Corcoran unpacked how the EBA’s ruling in G 1/23 quietly rattled European patent law. Dubbed the arrival of an “on-sale bar,” the decision expanded prior art to include any analysable feature of a marketed product, even if it can’t be reproduced. Corcoran argued this didn’t topple G 1/92 but subtly “recast” it, leaving the limits of undue burden undefined. Until another board draws that line, innovators would do well to treat every product they sell as an open book.Midway through their digital deep dive, the Kats stumbled on a true AI brain-twister: if certain AI systems are banned under the EU AI Act, does that mean they can’t be patented? As it turns out, not quite. Our Katfriends Andreas Engel and David Faber explained the Act’s prohibitions may sound like red lights, but with exceptions for research and development, and plenty of case-by-case nuance, innovation isn’t entirely boxed in. Regulation and patent law, it seems, can still coexist proving once again that even in a world ruled by algorithms, a little human reasoning goes a long way.
Geographical Indications
After all that screen time, Jocelyn Bosse gave the Kats something real to chew on with her review of The Future of Geographical Indications: European and Global Perspectives, edited by Andrea Zappalaglio and Enrico Bonadio. With forewords by Justin Hughes and Irene Calboli (a familiar name whose enthusiasm for GIs I still remember fondly), Jocelyn explored how GIs are adapting to global shifts in trade, culture, and climate. Her takeaway: while GIs promise sustainability and pride of place, authenticity still depends on context, not just origin.Update and Events
Finally, Jocelyn Bosse’s Friday Fantasies offered the Kats a well-earned break from all that screen time, rounding up the latest IP events and opportunities, for those ready to step out and meet the real world again.
Never Too Late: If you missed the IPKat last week!
Reviewed by Wissam Bentazar
on
Saturday, October 18, 2025
Rating:
Reviewed by Wissam Bentazar
on
Saturday, October 18, 2025
Rating:



No comments:
All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.
It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.
Learn more here: http://ipkitten.blogspot.com/p/want-to-complain.html