WAN-IFRA Media Policy BriefingTech wins two AI copyright cases; X rekindles ad biz using threats; More publishers sign with/sue AI companies; Tech investments in AI keep growing
Find news from WAN-IFRA Member Associations' community at the bottom of this email.
Three tech giants collected important judicial wins last week. A federal US judge ruled that the unauthorised use by Anthropic - backed by Amazon and Alphabet - of copyrighted books to train its Claude LLM was fair use. The same decision found that instead the copying and storage of 7 million pyrated books infringed copyright. A second judge concluded that Meta's use of books for training purposes without permission was transformative, and considering that the 13 authors who sued could not prove market harm the fair use test was met. The two decisions will likely inform ongoing and future content owners lawsuits, with strategies more focused on output rather than on input, and on proving the financial damage.
Liberal advocacy group Media Matters sued the Federal Trade Commission, saying that the FTC's investigation started against it was motivated by retaliation over negative reporting on the platforms X. The FTC suspects that Media Matters' reporting on X amounted to illegal collusion with advertisers. In the last few months X has succeeded in resuscitating the platform's advertising business by threatening lawsuits: the move brought back advertisers such as Verison, who had stopped advertising on X in 2022 and will now spend 10 USD millions this year. The technique has allowed the platforms to struck dels with at least another 6 major former advertisers. In the lawsuit of X against the World Federation of Advertisers, major companies filed for dismissal saying that X had lost their business on the free market and should not be allowed to use a courthouse to win it back.
More tran 500 publications licensed their content to GAI company ProRata, many following a framework agreement signed with WAN-IFRA member News/Media Alliance , who saluted the "innovative licensing and attribution system." Meanwhile the BBC has announced that it possesses evidence suggesting that Perplexisy's "default model" was "trained using BBC content". Legal action will be pursued unless the scraping stops and "proposal for financial compensation"is provided. Perplexity is already engaged in a legal battle with News Corp. For an updated list of who's signing with AI and who's suing check Pressgazette and Wired
AI startup Anthropic is launching Economic Futures, an effort to address the economic consequences of the rapidly advancing technology. The initiative follows recent comments of CEO Amodei who predicted that AI would eliminate within 5 years 50% of all entry-level white collar jobs and might handle almost all coding tasks within a year. Meta has instead announced a $14.3 billion investment in a "superintelligence hub", allegedly aimed most of all at attracting talented researchers with compensation packages up to $100 million. While return on the money remains elusive, Alphabet, Amazon, Meta and Microsoft plan to further accelerate their investments on AI, after increasing their capital spending on the technology by nearly two-thirds, or $95bn, in 2024.
Reply to this email to send us comments, questions or tips. Forward this message to anyone else who you think should be reading our media policy updates.
