Meredith Rose

us government has rendered it illegal to give access to fable or mythos to foreign nationals ^_^ www.anthropic.com/news/fable-m...

Statement on the US government directive to suspend access to Fable 5 and Mythos 5 anthropic.com View on Bluesky →

Tons of peoples' Claude chats are exposed on Google. Cryptocurrency keys, API keys, login credentials, legal discussions, more. This happened with ChatGPT last year. And Claude too. There's clearly an ongoing issue with search engines indexing peoples' AI chats www.404media.co…

Tons of Peoples’ Claude Chats and Creations are Exposed on Google 404media.co

AI Weekly's analysis →

• A missing noindex tag is the root cause, the same omission behind ChatGPT's 2025 indexing incident that exposed roughly 100,000 conversations.
• A GitHub repository archived 11,241 exposed messages before Anthropic's fix, making the breach permanent regardless of platform-side remediation.
• Anthropic's Google-side fix landed within hours, but Bing and Brave Search results persisted longer, exposing an uneven remediation window across search engines.

Read full analysis →

View on Bluesky →

"Whenever you get into the copyright realm, a lot of judges just decide it's all vibes." Senior Policy Counsel @mrose.ink breaks down Google and Reddit's uphill DMCA battles against @serpapi.com — and why the bigger risk is a broader "re-enclosure" of the open web. arstechnica…

“Google and Reddit do not own the Internet," web scraper says after court win arstechnica.com

AI Weekly's analysis →

• Chief U.S. District Judge Yvonne Gonzalez Rogers dismissed Google's DMCA suit against SerpApi on July 20, 2026, ruling public search results are not copyrighted.
• Google has 21 days to file an amended complaint, narrowed to Knowledge Panels where it says explicitly licensed media sits behind its SearchGuard anti-bot defense.
• SerpApi faces a parallel October 2025 Reddit lawsuit alleging it supplied user comments to Perplexity for AI training, and has moved to dismiss.

Read full analysis →

View on Bluesky →

Sex doll turned rural school district robot teacher is a reality now nysfocus.com/2026/07/14/n...

Rural NY School To Launch Humanoid Robot Teacher nysfocus.com View on Bluesky →

An AI-generated "news" slop site keeps making up EFF staffers and attributing made-up quotes to them in their stories: www.eff.org/deeplinks/20...

‘News’ Site Keeps Hallucinating EFF Staffers eff.org View on Bluesky →

The Agents of Young Performers Association has issued an open letter to Hasbro condemning them for forcing child actors on “Peppa Pig” to sign over the rights to their voice for reconstruction and use in generative AI to continue working on the show: deadline.com/2026/06/pepp...

'Peppa Pig' Backlash As Hasbro Asks Child Actors To Sign AI Clause deadline.com View on Bluesky →

Generative AI is trained on human writing. Programs that are supposed to detect AI writing do not work. There is no substantive evidence that they work. 50% false positive rate. May as well flip a coin. Substack is in the wrong b/c they are treating AI detection as valid. Sour…

Guides: Generative AI Detection Tools: The Problems with AI Detectors: False Positives and False Negatives lawlibguides.sandiego.edu View on Bluesky →

A timely reminder, as the judge's approval puts the Bartz settlement back in the news: the settlement doesn't tell us anything about AI and fair use. It's mostly proof that copyright statutory damages continue to create perverse incentives. recreatecoalition.org/dont-read-m...

Don’t Read Much Into the Proposed Settlement in Bartz v. Anthropic - Re:Create recreatecoalition.org View on Bluesky →