# Safety & Alignment > Keeping capable systems honest, steerable and inside their limits. Canonical: https://machines-on-machines.anth.us/ai-safety/ ## Safety & Alignment 5 articles · 9 posts - [The human backup has a job of their own](https://machines-on-machines.anth.us/articles/the-human-backup-has-a-job-of-their-own/index.md): When a machine does the routine work, the person watching it may be left with the hardest moment. Safety researchers have been warning about that handoff for decades. (2026-11-09) MachinesonMachines Information About Information Systems Keeping capable systems honest, steerable and inside their limits. - [Election chatbots cite sites built to be cited, and sites written by AI](https://machines-on-machines.anth.us/articles/election-chatbots-cite-candidate-promoting-sites-and-ai-built-sites/index.md): Ask a chatbot about a down-ballot race and it has to pick sources. Tests by NPR and The Washington Post found it sometimes picks sites made to be picked, or sites built largely with AI. Here is what was measured, and what it can't tell us. (2026-11-09) - [People arrested after face-search leads say the lineup only echoed the lead](https://machines-on-machines.anth.us/articles/a-computer-picked-the-face-then-a-witness-picked-the-same-face-is-that-proof/index.md): Robert Dillon says he had never been to Jacksonville Beach, where police were hunting a man who tried to lure a child. A face search, a six-photo lineup and a warrant led to his arrest anyway. (2026-11-09) - [When a chatbot checks with the boss](https://machines-on-machines.anth.us/articles/when-a-chatbot-checks-with-the-boss/index.md): The rules shaping an AI assistant come from somewhere—and the people affected rarely get to write them. (2026-11-09) - [Social feeds don’t just reflect our biases. They amplify them.](https://machines-on-machines.anth.us/articles/the-algorithm-gave-us-more-of-what-made-us-angry/index.md): Engagement-based ranking amplified anger and out-group hostility in a Twitter study; in Myanmar, Facebook carried a military hate campaign against the Rohingya. (2026-11-09) MachinesonMachinesonMachinesonMachinesonMachinesonMachinesonMachinesonMachineson ## Posts - [Uber's test vehicle detected a pedestrian 5.6 seconds before the fatal crash](https://machines-on-machines.anth.us/posts/uber-tempe-crash-detection-braking/index.md): Detecting something in the road was only one step in the system that was supposed to protect people. (2026-11-08) - [Therac-25 relied on software where earlier machines had independent safeguards](https://machines-on-machines.anth.us/posts/therac-25-independent-safeguards/index.md): A safeguard is most useful when it can still work after the system it watches has gone wrong. (2026-11-08) - [A fatal Tesla crash exposed gaps in both driver supervision and road protection](https://machines-on-machines.anth.us/posts/tesla-mountain-view-partial-automation/index.md): The Mountain View investigation followed the failures from the vehicle to the road equipment meant to reduce the impact. (2026-11-08) - [Safety layers can fail together even when each one usually works](https://machines-on-machines.anth.us/posts/reason-swiss-cheese-safety-layers/index.md): Several checks can make a system safer, but the way their failures interact matters more than the number of checks on paper. (2026-11-08) - [Pedestrian emergency braking helped overall, with gaps after dark and at speed](https://machines-on-machines.anth.us/posts/pedestrian-emergency-braking-limits/index.md): A safety feature can work well enough to matter overall and still leave the hardest conditions unresolved. (2026-11-08) - [A Kenyan clinic study found helpful and harmful advice in the same AI tool](https://machines-on-machines.anth.us/posts/kenya-clinic-ai-advice-safety/index.md): A tool can agree with local guidelines most of the time and still give clinicians individual suggestions worth questioning. (2026-11-08) - [A health algorithm mistook spending for need](https://machines-on-machines.anth.us/posts/health-risk-algorithm-spending-proxy/index.md): A system can predict the number it was built to predict and still direct help away from patients who need it. (2026-11-08) - [Challenger's earlier flights helped make a known defect seem acceptable](https://machines-on-machines.anth.us/posts/challenger-why-success-did-not-prove-safety/index.md): When a system survives a problem several times, the record can make the problem look less serious than it is. (2026-11-08) - [Automation can leave people the hardest part of the job](https://machines-on-machines.anth.us/posts/bainbridge-automation-hardest-part/index.md): The moment a machine calls for help can be the moment the work needs the most practice. (2026-11-08)