[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-models-are-stricter-on-fairness-than-humans-study-finds":10,"sections":35},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":24,"tags":25,"sources":30,"feedback":34,"feedback_at":22,"cost_usd":34,"total_tokens":34},7718,"ai-models-are-stricter-on-fairness-than-humans-study-finds","AI Models Are Stricter on Fairness Than Humans, Study Finds","New research comparing LLM and human fairness judgments finds models favor stricter rules, act more self-interested, and resist fine-tuning toward human norms.","Ask an AI to divide up scarce resources, and it turns out to be pickier about fairness than you are.\n\nResearchers built a general method to test how large language models reason about distributional justice - who gets what when there isn't enough to go around. They ran matched scenarios past a broad set of LLMs and human subjects, using identical framing and elicitation conditions for both groups. The models consistently preferred stricter fairness constraints than people did. They also showed more self-interested behavior, shifted their answers depending on how a scenario was worded, and stayed hard to align with human judgments even after fine-tuning on current datasets.\n\nThat last finding is the one that should worry deployers. Companies already lean on LLMs to help allocate loans, benefits, seats, and other limited resources, often assuming a fine-tuning pass can nudge a model toward whatever fairness standard the business wants. This research suggests that assumption doesn't hold: existing alignment techniques and datasets aren't reliable enough to steer a model's sense of fair play.\n\nNo fairness theory is bulletproof - philosophers have argued about the right one for decades. But humans at least know they're guessing. An LLM hands back a confident, rigid answer regardless, and confident wrongness is a harder problem to catch than an honest shrug.","[\"ai alignment\",\"fairness\",\"llm research\",\"ai ethics\"]","2026-09-25T04:00:00.000Z","2026-09-25T06:12:36.719Z","2026-09-25T06:12:42.049Z","published",null,[],"ai",[26,27,28,29],"ai alignment","fairness","llm research","ai ethics",[31],{"name":32,"url":33},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.29692",0,{"sections":36},[37,40,45,50,55,60,65,70,75,80,85,90,94,99],{"name":38,"slug":24,"count":39,"latest_published_at":18},"AI",4466,{"name":41,"slug":42,"count":43,"latest_published_at":44},"Security","security",729,"2026-09-24T19:54:21.000Z",{"name":46,"slug":47,"count":48,"latest_published_at":49},"Policy","policy",386,"2026-09-24T23:50:55.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Deals","deals",237,"2026-09-24T22:00:00.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",182,"2026-09-25T01:25:53.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Science","science",138,"2026-09-24T18:24:52.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Consumer Tech","consumer-tech",128,"2026-09-24T19:24:34.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Software","software",88,"2026-09-24T23:06:55.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Dev Tools","dev-tools",79,"2026-09-22T22:21:13.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Startups","startups",71,"2026-09-24T20:45:00.000Z",{"name":86,"slug":87,"count":88,"latest_published_at":89},"Gaming","gaming",46,"2026-09-24T17:52:29.000Z",{"name":91,"slug":92,"count":88,"latest_published_at":93},"General","general","2026-09-25T02:12:57.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",30,"2026-09-24T20:07:31.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]