Amazon Atlas Union Risk AI & AI Abolition Manifesto Gains Attention - AI News (Oct 18, 2026)

Support The Automated Daily directly: 15 free episodes a day, 5 editions, 3 languages, about $4 a day to produce. A coffee keeps it going: theautomateddaily.com/support Please support this podcast by checking out our sponsors: - SurveyMonkey, Using AI to surface insights faster and reduce manual analysis time - https://get.surveymonkey.com/tad - Prezi: Create AI presentations fast - https://try.prezi.com/automated_daily - Discover the Future of AI Audio with ElevenLabs - https://try.elevenlabs.io/tad Today's topics: Amazon Atlas Union Risk AI - An openDemocracy investigation reports that Amazon uses an internal AI tool called Atlas to score warehouses in the US, Canada and the UK on their likelihood of unionising. The findings raise concerns about workplace surveillance and labor rights, and Amazon says the tool works only at the site level. AI Abolition Manifesto Gains Attention - A new manifesto argues that AI is tied to capitalism and control and calls for AI abolition rather than harm reduction. It cites environmental damage, labor exploitation, privacy loss and authoritarianism, and it urges organizing and less reliance on AI tools. Caseway Rejects US Government Work - Vancouver AI firm Caseway is turning down US government opportunities because of Canada-US tensions and Trump's '51st state' remarks. It will focus on Canadian defence work and on customers in Europe and Taiwan, a choice that shows how dependent Canada's AI sector is on US investment. HUMXN Trades Repairs for Robot Data - AI lab HUMXN offers homeowners in Minneapolis, Chicago and Miami free plumbing, electrical and HVAC repairs if technicians wear capture gear during the job. The recordings will become robotics training data, a sign that AI companies are starting to pay for physical-world data. Verification and Comprehension Checks for AI Code - As AI coding agents speed up software development, a new article calls for verification as a core part of CI/CD, combining tests, security scans and AI code review. Separately, the open-source Git hook Sphinx asks developers questions about agent-written code before it lets a push go through. Behavioral Testing Exposes AI Agent Bugs - Behavioral testing of 53 open-source AI agents turned up 37 bugs across about 3,000 scenarios. The failures included unsafe file deletion without approval, which points to gaps in agent permissions and safety testing that text-only evaluations miss. - Amazon Accused of Using AI to Track Union Risk in UK Warehouses