Bundesrat Demands Wider AI Surveillance. US Blocks Fable and Mythos. Count Anything Improves Visual Counting. Gemini-SQL2 Translates Language to SQL.

Show notes

The AI news for June 14th, 2026

--- This episode is sponsored by ---

Rocket Routine GmbH
Find our more about our today's sponsor Rocket Routine at
The Federal Council wants AI surveillance and a Quick Freeze by the BKA.
Source: https://www.heise.de/news/Biometrische-Rasterfahndung-Bundesrat-will-noch-schaerfere-KI-Ueberwachung-11331325.html?wt_mc=rss.red.ho.themen.k%C3%BCnstliche+intelligenz.beitrag.beitrag
Why did we choose this article?
A push by German regional governments to loosen rules for automated biometric matching and 'quick freeze' investigations would expand state surveillance power and directly affect citizens' privacy, businesses operating in Germany, and online services.

The U.S. government forces the shutdown of Anthropics AI Fable 5 and Mythos 5.
Source: https://www.heise.de/news/US-Regierung-erzwingt-Abschaltung-von-Anthropics-KI-Fable-5-und-Mythos-5-11331129.html?wt_mc=rss.red.ho.themen.k%C3%BCnstliche+intelligenz.beitrag.beitrag
Why did we choose this article?
A US export directive forcing Anthropic to shut down two widely discussed models removes user and business access overnight, showing how geopolitics and regulation can immediately disrupt AI services people rely on.

AI model 'Count Anything' solves a problem that multimodal AI models have so far failed to solve.
Source: https://the-decoder.de/ki-modell-count-anything-loest-ein-problem-an-dem-multimodale-ki-modelle-bisher-scheitern/
Why did we choose this article?
A model that can count objects in arbitrary images from a text prompt can automate tasks across events, retail, safety monitoring and microscopy, changing how organizations measure crowds, inventory, or cell samples without custom tooling.

Google's Gemini-SQL2 translates natural language into SQL queries.
Source: https://the-decoder.de/googles-gemini-sql2-uebersetzt-natuerliche-sprache-in-sql-abfragen/
Why did we choose this article?
Turning natural-language requests into executable SQL makes databases accessible to non-technical knowledge workers, letting teams pull actionable data without hiring SQL experts or writing queries by hand.

Do you have any questions, comments, or suggestions for improvement? We welcome your feedback at podcast@pickert.de.