Bartz v. Anthropic — Final Approval Approaching #
- Bartz v. Anthropic Settlement: What Authors Need to Know (Authors Guild) — Final approval hearing: May 14, 2026, 2 p.m. PT. $1.5B total settlement; ~$3,000 per title (may increase based on claims submitted). Covers ~500,000 book titles downloaded from LibGen (June 2021) and PiLiMi (July 2022). Claims deadline passed March 30, 2026. 50/50 author/publisher split for trade books by default; self-published authors receive full award.
AI Output: Discovery Orders & Authorship Ruling #
Cross-links #
- AI Impact on Society The $3,000/title Bartz payout establishes a pricing floor for AI training-data licensing — watch whether this becomes the benchmark for future licensing deals (as UMG/Udio established for music).
- Open vs Closed AI Ecosystems Output-log discovery orders apply to closed labs (OpenAI, Anthropic) because they control and retain logs. Open-weight models without centralised inference are structurally less exposed to this discovery mechanism.
- Claude-Specific Expertise Bartz final approval (May 14) removes one major litigation uncertainty for Anthropic — watch for any impact on Managed Agents commercial expansion timing.
- Emerging theme: Output-log discovery is the new litigation frontier — courts are using log production orders to test whether AI outputs reproduce training material, making output-level infringement claims empirically testable for the first time.
- Emerging pattern: The Bartz settlement structure (~$3,000/title, piracy-pathway liability) is becoming the template for future settlements. The music publishers’ $3.1B ask is calibrated against this floor; watch the per-composition calculation in that case.
- Keyword suggestion: “output-log discovery” — the mechanism courts are using to operationalise output infringement claims; distinct from training-data fair-use analysis.
- Quality signal: Taylor Wessing’s analysis of the Supreme Court certiorari denial is the clearest statement that AI-generated output remains uncopyrightable under US law regardless of human prompting — important for IP strategy.