AI images fooling humans & Human authorship and trust - AI News (Sep 20, 2026)

Please support this podcast by checking out our sponsors: - Invest Like the Pros with StockMVP - https://www.stock-mvp.com/?via=ron - KrispCall: Agentic Cloud Telephony - https://try.krispcall.com/tad - Discover the Future of AI Audio with ElevenLabs - https://try.elevenlabs.io/tad Support The Automated Daily directly: Buy me a coffee: https://buymeacoffee.com/theautomateddaily Today's topics: AI images fooling humans - A fast image quiz called Reality Check shows how difficult it has become to tell real photos from AI-generated pictures. The story highlights synthetic media, image realism, and the growing challenge of visual trust online. Human authorship and trust - A new essay argues people should rarely use AI to draft substantive writing because writing is part of thinking, and unlabeled AI prose can erode reader trust. It puts human authorship, reasoning, and disclosure at the center of the AI debate. Open source AI backlash - Another commentary warns generative AI is weakening the culture of sharing that helped build modern software, while KDE's proposed AI-native desktop shows how divided open-source communities have become. Keywords here are open source, scraping, licensing, and AI-native computing. NYT copyright case update - In a new development, The New York Times says internal documents strengthen its copyright claims against OpenAI and Microsoft. The case could shape AI training data rules, fair use arguments, and publisher economics. Antitrust fight over slowdown - A fresh lawsuit now claims major AI labs may have crossed an antitrust line by publicly aligning around slower AI development. The dispute could define how companies discuss AI safety standards without appearing to coordinate against competition. Google AI deletion controversy - A researcher claims Google AI Studio's deletion flow does not fully erase chat data, raising privacy and compliance concerns. The allegation also draws attention to data retention, user expectations, and bug bounty handling. - Why the Author Says You Should Almost Never Use AI to Write - NYT Lawsuit Briefs Reveal Harsh Internal Warnings About AI’s Impact on Publishers - Reality Check Game Tests Whether Images Are AI or Real - Bluesky Post Critiques AI Safety Community - AI Is Undermining the Open-Source Commons - Lawsuit Says Major AI Firms Illegally Coordinated a Slowdown ([apnews.com](https://apnews.com/article/antitrust-lawsuit-ai-slowdown-anthropic-openai-spacexai-google-960af4308161eaf4ed13c383b0ce1c1b)) - Author Alleges Google AI Studio Delete Button Does Not Really Delete Data - Insiders Say OpenAI and Anthropic Oversold AI Breach Fears to Sway Regulators - KDE’s 30th Anniversary Meets an AI-Native Desktop Debate Episode Transcript AI images fooling humans Let's start with the visual side of AI. A quick game called Reality Check asks people to decide whether images are real photos or AI-generated, and the point is not really the scoring. The real takeaway is that synthetic images are now convincing enough that many people will hesitate, second-guess themselves, or simply get it wrong. That matters well beyond entertainment, because the harder it is to spot fake visuals, the easier it becomes for misinformation, scams, and low-trust content to blend into everyday browsing. Human authorship and trust Two pieces today push back on a bigger cultural shift. One argues that people should almost never use AI to write substantive text, not because the output is always terrible, but because writing is part of thinking. If you hand that work to a model too early, you may skip the hard part of testing your own argument. A related essay says the same kind of shortcut is putting pressure on the open internet itself, as AI systems absorb shared writing and code without clearly honoring the social norms and licenses that made openness productive in the first place. Put together, the message is simple: AI can save time, but it can also weaken judgment and trust if it replaces the work rather than supporting it. Open source AI backlash That broader tension is showing up in open-source communities too. KDE, which is marking its 30th anniversary, is already seeing debate around a proposed AI-native desktop for Plasma. The idea is to make AI part of the core computing experience rather than just another assistant window, and that is exactly the sort of shift that divides technical communities right now. For some, it sounds like the next interface. For others, it sounds like a loss of control, privacy, and simplicity. Either way, it shows that AI is no longer just a tool discussion. It is becoming a product philosophy discussion. NYT copyright case update In a new development in the copyright fight we've been following, The New York Times is seeking summary judgment against OpenAI and Microsoft, saying newly cited internal documents show executives understood the risk AI systems posed to publishers. According to the reporting, the filings include blunt internal comments