The day's verified AI news for 2026-07-20. Every claim checked against the primary source.

Judge Grants Final Approval to Anthropic's $1.5 Billion Book-Piracy Settlement

A federal judge granted final approval of Anthropic's $1.5 billion class settlement with authors, entered judgment, and ordered the pirated book files destroyed within 30 days.

Read on Ground Truth · primary source

Axios: U.S. Officials Revive an Effort to Discourage Chinese Open-Weight AI

Axios reports that internal U.S. efforts to restrict Chinese open-weight AI models have revived after Kimi's rise, but no ban, rule, or executive order has been announced.

Read on Ground Truth · primary source

A Mathematician Posts a Counterexample to a Famous Conjecture, Crediting an AI Model

Mathematician Levent Alpoge posted a hand-checkable counterexample to the Jacobian conjecture and credited the AI model Fable; the math is independently auditable, but the AI's actual role is not documented.

Read on Ground Truth · primary source

Safety Guardrails Blocked a Security Team's Own Incident Analysis

Hugging Face disclosed that commercial AI safety filters blocked its analysis of real attack code during an incident, so it ran the forensics on a self-hosted open-weight model instead.

Read on Ground Truth · primary source

Google Falls Off One Leaderboard's Top 15, as a Report Describes a Gemini-Specific Chip

Google dropped out of the top 15 on LLM Stats' composite leaderboard while remaining its fastest model, and Reuters separately reported an unannounced Gemini-specific inference chip.

Read on Ground Truth · primary source

VideoChat3 Halves Video-Model Latency by Compressing Space and Time First

VideoChat3, an open 4-billion-parameter video model, compresses frames across space and time before the language model, roughly halving latency versus a comparable model.

Read on Ground Truth · primary source

Tencent's Open Robot Model Plans by Imagining the Scene It Wants to Create

Tencent released RxBrain, an open ~6.2B robot model that interleaves text reasoning with generated goal images, betting that a robot needs an explicit picture of the world it is trying to build.

Read on Ground Truth · primary source

Microsoft's Resource2Skill Compiles Tutorials and Repos Into Executable Agent Skills

Microsoft released Resource2Skill, a runtime that turns tutorials, code repositories, and articles into structured, executable agent skills, reporting an average 11.9-point performance lift.

Read on Ground Truth · primary source

The Director of the U.S. AI-Evaluation Agency Is Leaving After Three Months

CAISI Director Chris Fall is leaving after about three months, with NIST Director Arvind Raman becoming acting head, days after the agency published a detailed assessment of a Chinese open-weight model.

Read on Ground Truth · primary source

You are getting this because you subscribed at groundtruth.day.