Judge Approves Anthropic's $1.5B Copyright Deal as Meta Weighs a $10B Bet

A legal liability becomes a fixed cost

The $1.5 billion settlement Anthropic reached over the use of copyrighted material in model training received formal judicial sign-off, converting what had been an open-ended legal exposure into a quantified, closed obligation. For a company preparing for large capital commitments elsewhere, as the Meta talks reported the same day illustrate, removing this uncertainty from the balance sheet has value beyond the settlement figure itself.

The approval also sets a reference point for the wider industry: it is one of the largest publicly disclosed resolutions of an AI copyright dispute, and its terms are likely to be cited in other pending litigation involving training-data provenance.

Sources: Un juge américain approuve l'accord à l'amiable conclu par Anthropic dans le cadre d'un litige en matière de droits d'auteur, d'un montant de 1,5 milliard de dollars · Un juge américain valide l'accord de 1,5 milliard de dollars d'Anthropic pour mettre fin aux poursuites sur le droit d'auteur · Un juge américain approuve l'accord à l'amiable d'Anthropic de 1,5 milliard de dollars dans le cadre d'un liti

Meta circles a multi-billion-dollar arrangement

Separately, several outlets reported that Meta is negotiating a deal with Anthropic valued at roughly $10 billion, with at least one report specifying that the contract concerns AI computing capacity. Coming from a company Anthropic competes with directly at the model layer, a commercial arrangement of this size would signal that compute economics — securing capacity, spreading infrastructure costs, or monetizing surplus capability — are pushing frontier labs toward deals with rivals even as they continue to race on capability.

None of the reports confirm final terms, and the figure is described as still under negotiation rather than signed. But the scale discussed places it among the larger disclosed inter-lab commercial arrangements to date, and it lands the same week Anthropic closed out a $1.5 billion legal liability — a juxtaposition that underscores how much capital is now moving through the sector on both the legal and commercial sides.

Sources: Meta et Anthropic discuteraient d'un accord à 10 milliards de dollars · Meta prépare un énorme accord à 10 milliards de dollars avec Anthropic · Meta négocie avec Anthropic un contrat de plusieurs milliards pour de la puissance de calcul IA

Fable 5's competitive reception sours

On the product side, the day brought a cluster of critical signals around Claude Fable 5. One report described Anthropic delivering unwelcome news to Fable 5 users, while a separate piece of commentary argued that Fable 5's output style reads closer to Moonshot's Kimi K3 than to Anthropic's own Claude Opus — a comparison that, if it holds, cuts against the idea of a coherent Claude 'house style' across model generations.

The same day, Alibaba pitched its newly unveiled Qwen3.8 as a rival to Fable 5, and a separate analysis framed the emergence of Kimi K3 and Qwen3.8 as a test of frontier-lab economics, questioning Anthropic's competitive footing. None of this amounts to confirmed benchmark results, but the volume and timing of the commentary — arriving alongside the Meta compute talks — suggests Anthropic is fighting a two-front battle: securing capital and capacity even as rival labs in China press on model quality and pricing.

Sources: Anthropic annonce une mauvaise nouvelle aux utilisateurs de Claude Fable 5 · Claude Fable is stylistically closer to Kimi K3 than Claude Opus · Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling · Alibaba présente Qwen3.8, affirme qu'il rivalise avec Anthropic Fable 5

The rivalry spills into public messaging

Beyond product and legal news, the day's items also captured competition playing out in the open: Sam Altman publicly mocked Anthropic's latest advertising campaign, which some online commentators described as 'dystopian marketing.' Whatever the merits of the ad itself, the exchange is a reminder that as the leading labs converge on similar capabilities and now, per the Meta reports, similar commercial partners, differentiation is increasingly being fought over brand perception as much as benchmarks.

Sources: Sam Altman raille la dernière publicité d'Anthropic, qualifiée de « marketing dystopique » par les internautes