Summary
Overview
The WAN Show discusses major AI developments including Google's liability for false AI Overview information, Meta's pulled facial recognition system, and Anthropic's Fable 5 shutdown. Other topics include Florida hospital using Palantir to detect sepsis, Instagram's new algorithm controls, donut battery fraud exposure, Canada's proposed social media ban for under-16s, and various tech and gaming news.
Google Found Liable for AI Overview Misinformation
A German court ruled Google liable for false information from AI Overview after publishing companies sued when Google's AI combined their information with shady sources and fabricated data not found in citations. Google plans to appeal, arguing users can fact-check themselves and claiming these are narrow errors. The hosts demonstrate how AI Overview provides inconsistent results and often cites sources that don't contain the quoted information, with one test showing the disclaimer about AI making mistakes only appearing inconsistently across browsers.
- German Regional Court of Munich ruled Google liable for false AI Overview information
- Google's AI combined legitimate publishers with shady companies and drew information not in linked sources
- Testing showed AI Overview provides different results for same query and often lacks the disclaimer about potential errors
- Google plans to appeal, claiming users can fact-check and these are narrow, not foundational errors
" users can fact-check for themselves, and they do plan to appeal the ruling with a spokesperson saying, this case focuses on specific and narrow errors, not the foundational way that AI overviews display web content. "
" If we all know, right? It's kind of like what went down with end-user license agreements, right? Like you had these long, theoretically legally binding documents. Everybody knew on both sides, the companies writing these end-user license agreements, the people signing them. Everybody knew that nobody was reading them. "
Florida Hospital Uses Palantir to Save Lives
Tampa General Hospital has used Palantir's Sepsis Hub to save an estimated 886 lives since 2022 by monitoring real-time data from 1,000 patients simultaneously and alerting rapid response teams. The system cut sepsis deaths and reduced hospital stays by 30%. However, this raises privacy concerns as Palantir receives health records, lab results, and bedside monitor data from all patients, aligning with the company's broader surveillance goals despite the positive medical outcomes.
- System monitors approximately 1,000 patients at once using real-time electronic health records, lab results, clinician notes, and bedside monitors
- Sepsis Hub has saved estimated 886 lives since 2022 and cut hospital stays by 30%
- Sepsis kills around 350,000 Americans yearly, making it the leading cause of death in US hospitals
- Palantir publicly advocates for total surveillance with no privacy, making their access to all this medical data concerning
" Are you okay with the health records, lab results, clinician notes, bedside monitor access, and probably tons of other things being mass data harvested from your loved one. Oh, also, they might not die if you click this checkbox. "
" looking at Palantir's playbook, they want to own all of the data in the world. They've basically publicly said that they don't want people to have any form of freedom and that people won't commit crimes anymore when there is genuinely no such thing as privacy "
📚 6 more sections below
Sign up to unlock the complete summary with all insights, key points, and quotes
Get this summary + all future The WAN Show episodes in your inbox
100% Free • Unsubscribe Anytime
Sign up now and we'll send you the complete summary of this episode, plus get notified when new The WAN Show episodes are released—delivered straight to your inbox within minutes.