WikiHow has sued OpenAI for using its content without permission in AI training. What makes this case legally distinct from *The New York Times* or Getty Images suits is that WikiHow published its content under Creative Commons licensing—free to use, but not for commercial purposes. That NC clause may be the cleanest copyright argument against training data use yet assembled.
Aug 25, 2026 at 16:27 7 8
Netflix paid $587M to acquire the AI startup InterPositive. A clear signal about the place of AI in the next Hollywood cost cycle.
Jul 20, 2026 at 11:40 9 12
French publishers have formally asked France's competition regulator to act against Google's AI news summaries. It's the sharpest test yet of whether EU neighbouring rights apply to the AI inference layer.
Aug 14, 2026 at 18:59 6 7
Apple is negotiating direct licensing deals with publishers to supply Siri with current news content—a signal that fresh, grounded news is genuinely differentiated from what AI models can produce from training data alone.
Aug 13, 2026 at 20:43 9 11
Three architectures to dissect side-by-side: Poolside's MoE 118B / 8B active on a single machine; Inkling's multimodal 975B / 41B, also released as 276B / 12B; and Kimi K3's 2.8T / 104B context with 1M tokens, whose commercial agreement clause could exclude US companies.
Aug 3, 2026 at 00:53 16 13
Mozilla launched on July 17 a state-of-the-art website on open source AI - an anchor point at a time when the open model economy is faltering.
Jul 17, 2026 at 22:03 26 9
The first open-weights model from Thinking Machines Lab (Mira Murati) is here: Inkling, MoE 975B params (41B active), multimodal, Apache-2.0 license. The most discreet box in the frontier scene enters the open market through the front door.
Jul 16, 2026 at 21:46 25 8
