Tax Offices Train on Real Data. Anthropic Bans AI in Interviews. SoftBank Builds French AI Hubs. Search Agents Lean on Memory.

Show notes

The AI news for June 1st, 2026

--- This episode is sponsored by ---

Rocket Routine GmbH
Find our more about our today's sponsor Rocket Routine at
Tax Office 2.0: Tax authorities should train AI with real citizen data.
Source:
https://www.heise.de/news/Finanzamt-2-0-Steuerbehoerden-sollen-KI-mit-echten-Buergerdaten-trainieren-11312693.html?wt_mc=rss.red.ho.themen.k%C3%BCnstliche+intelligenz.beitrag.beitrag
Why did we choose this article?
This proposes letting tax authorities train AI on real citizen data (with a one-year deletion limit) — a direct change to privacy, oversight and how public services will use personal data that affects taxpayers and businesses.

Anthropic prohibits the use of AI in the job interview.
Source: https://the-decoder.de/anthropic-verbietet-ki-nutzung-im-bewerbungsgespraech/
Why did we choose this article?
A major AI company is banning candidates from using AI during interviews and changing its hiring process — this alters expectations for job-seekers, interview prep, and the market for interview coaching.

Up to €75 billion: SoftBank makes France the largest European location for AI infrastructure.
Source: https://the-decoder.de/bis-zu-75-milliarden-euro-softbank-macht-frankreich-zum-groessten-europaeischen-standort-fuer-ki-infrastruktur/
Why did we choose this article?
A potential €45–75B plan for large AI data centers in France could shift where European AI infrastructure, jobs and energy use concentrate — affecting cloud choice, local economies and policy debates even before projects are built.

AI search agents often 'google' only what they already know anyway.
Source: https://the-decoder.de/ki-suchagenten-googeln-oft-nur-was-sie-ohnehin-schon-wissen/
Why did we choose this article?
Research shows popular AI search agents often don't reliably fetch fresh, time-sensitive facts and instead lean on learned knowledge — important for anyone relying on AI for up-to-date research, news, compliance or business decisions.

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

New comment

Your name or nickname, will be shown publicly
At least 10 characters long
By submitting your comment you agree that the content of the field "Name or nickname" will be stored and shown publicly next to your comment. Using your real name is optional.