Anthropic

In the publisher-licensing market, Anthropic is the subject of what is currently the largest single publicly named AI content deal: News Corp chief executive Robert Thomson disclosed on 8 May 2026 that News Corp anticipates a $1.5 billion settlement with Anthropic -- an 'anticipated' figure read into an SEC-filed earnings transcript, not yet a closed agreement. Judge Araceli Martinez-Olguin took the Bartz v Anthropic class settlement under submission at the 14 May fairness hearing, with a claims rate of 92.77% (447,576 of 482,460 works claimed) and a fee petition of $187.5 million; the judge subsequently ordered supplemental briefing on late opt-outs. Final approval has not been entered.

The Center for Journalism and Liberty's April 2026 report identified Anthropic's scrape ratio at 73,000 pages per human visitor returned, versus 1,700:1 for OpenAI -- a ratio that shapes Anthropic's relative exposure in publisher litigation and explains why its settlement figure exceeds News Corp's previously disclosed five-year OpenAI deal by roughly six times. Akamai signed a $1.8 billion, seven-year compute supply deal with Anthropic on 8 May 2026 -- the largest contract in Akamai's history -- placing Claude's inference at the EU network edge.

On 12 June 2026, the US government did something it had never done to a consumer technology product: Commerce Secretary Howard Lutnick sent Anthropic chief executive Dario Amodei a letter ordering the withdrawal of Claude Fable 5 and Claude Mythos 5 from all foreign nationals, inside and outside the United States, including Anthropic's own foreign-national staff. To comply, Anthropic disabled both models for every customer on earth. The trigger was a rival company's claimed jailbreak of Mythos 5. Anthropic publicly disagreed, noting the same vulnerability sits in GPT-5.5, which remained on sale. The administration invoked national-security authority without naming a statute.

The political context sharpens the episode. Anthropic had backed the pro-regulation Public First campaign; the OpenAI camp, through Greg Brockman and Andreessen Horowitz, funded the anti-regulation Leading the Future PAC. The lab that lobbied to be regulated was the one the administration pulled first. The suspension sets a precedent: runtime access to a consumer AI model can be treated as a controlled export under the deemed-export doctrine built for chips and closed weights. The US Commerce Department partially lifted the restriction on 27 June 2026, fifteen days after the original order, converting the hard cut-off into an expedited review process.

The Trump administration's June 2026 order against Anthropic's top models arrived at the worst possible moment for EU AI strategy. The EU Agency for Cybersecurity (ENISA), admitted to Anthropic's Project Glasswing restricted government-access programme in April 2026, lost access to Fable 5 and Mythos 5 within weeks of joining. MEP Christophe Grudler described the episode as a 'real kill-switch', arguing that European governments depending on US frontier models faced revocability risk no self-hosted system carries.

The US Commerce Department partially lifted the restriction on 27 June 2026, five weeks before the EU AI Act's GPAI enforcement rules activate on 2 August 2026. The partial reversal came fifteen days after the original order but did not eliminate the precedent. The episode hands European alternatives, chiefly Mistral AI and Aleph Alpha, their strongest commercial argument: ENISA's experience shows concretely that relying on a US-controlled frontier model for government cybersecurity functions carries a political-risk dimension that a self-hosted European alternative does not.