Apple Intelligence Repeats Stereotypes. Voice AI Echoes False Claims. MPA Moves Against Seedance. AI Agents Stick to Coding.

Show notes

The AI news for February 23rd, 2026

--- This episode is sponsored by ---

Pickert GmbH
Find our more about our today's sponsor Pickert at pickert.de.

---

Would you like to create your own AI-generated and 100% automated podcast on your chosen topic? --> Reach out to us, and we’ll make it happen.

Here are the details of the day's selected top stories:

Apple Intelligence hallucinates stereotypes in millions of summaries.
Source: https://the-decoder.de/apple-intelligence-halluziniert-stereotype-in-millionen-von-zusammenfassungen
Why did we choose this article?
An independent audit finds bias in automatic summaries on hundreds of millions of Apple devices — this directly affects users' inbox overviews, could misrepresent people or situations, and has implications for privacy, fairness, and whether you should trust or opt out of automatic summaries.

Disinformation via audio: ChatGPT Voice and Gemini Live can be easily manipulated.
Source: https://the-decoder.de/desinformation-per-audio-chatgpt-voice-und-gemini-live-lassen-sich-leicht-manipulieren/
Why did we choose this article?
Voice-capable AI assistants can be tricked into producing realistic but false audio 'news' — a direct risk for misinformation, scams, and trust in spoken information that consumers and news organizations need to watch for.

AI video generator Seedance 2.0: Hollywood's industry association sends a cease-and-desist letter to ByteDance for the first time.
Source: https://the-decoder.de/hollywoods-branchenverband-schickt-erstmals-unterlassungsschreiben-an-ki-unternehmen/
Why did we choose this article?
Hollywood's trade group has taken legal action against an AI video generator, signaling real legal and licensing risks for AI-generated film/video content — important for creators, platforms, and anyone using such tools commercially or creatively.

AI agents, according to Anthropic, have so far only somewhat gained traction in coding.
Source: https://the-decoder.de/ki-agenten-sind-laut-anthropic-bislang-nur-im-coding-einigermassen-angekommen/
Why did we choose this article?
Vendor data suggests AI 'agents' are mainly useful today for coding, not yet a broad substitute for human knowledge work — a practical signal for managers to pilot agents in specific tasks rather than expect wholesale automation.

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