[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-ai-content-filters-still-fail-when-the-rules-change":10,"sections":34},{"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":29,"feedback":33,"feedback_at":22,"cost_usd":33,"total_tokens":33},4364,"ai-content-filters-still-fail-when-the-rules-change","AI Content Filters Still Fail When the Rules Change","A new benchmark exposes how image safety models break down the moment a platform updates its content policy, and proposes a fix.","Most AI image filters assume safety is baked into a picture. It is not.\n\nResearchers introduced PolicyShiftBench, a benchmark of 2,000 test instances across 265 images, designed to measure whether guardrail models can adapt to a policy that changes mid-deployment — not just a fixed global ruleset. The same image might be acceptable on a general platform, blocked on a children's service, and newly restricted after a policy revision. Existing vision-language models and commercial guardrails, it turns out, largely ignore those distinctions and keep leaning on their trained-in safety priors. The team also proposed PolicyShiftGuard, a 7-billion-parameter model trained with a two-stage recipe that pairs matching allow\u002Fblock examples for the same image to teach the model the difference between a blocking policy and a passing one. On the new benchmark, it reached 76.9 average F1 — currently the best reported number on PolicyShiftBench.\n\nThe gap this exposes is real and underappreciated. Every major platform adjusts content policies over time, and retraining a guardrail model from scratch each time is expensive and slow. A policy-conditioned model that can generalize to new rules without retraining would cut that cycle considerably. The researchers also report that PolicyShiftGuard transfers reasonably well to two external benchmarks — UnSafeBench and SafeEditBench — which suggests the gains are not just benchmark-specific tuning.\n\nA 76.9 F1 on a benchmark built by the same team that built the model is a number worth treating carefully until independent replication arrives.","[\"ai\",\"content moderation\",\"safety\",\"research\"]","2026-07-08T04:00:00.000Z","2026-07-08T07:14:02.873Z","2026-07-08T07:14:05.846Z","published",null,[],"ai",[24,26,27,28],"content moderation","safety","research",[30],{"name":31,"url":32},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.05910",0,{"sections":35},[36,40,45,50,55,60,65,70,75,80,85,89,94,99],{"name":37,"slug":24,"count":38,"latest_published_at":39},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":41,"slug":42,"count":43,"latest_published_at":44},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":46,"slug":47,"count":48,"latest_published_at":49},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":51,"slug":52,"count":53,"latest_published_at":54},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":56,"slug":57,"count":58,"latest_published_at":59},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":61,"slug":62,"count":63,"latest_published_at":64},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":66,"slug":67,"count":68,"latest_published_at":69},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":71,"slug":72,"count":73,"latest_published_at":74},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":76,"slug":77,"count":78,"latest_published_at":79},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":81,"slug":82,"count":83,"latest_published_at":84},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":86,"slug":87,"count":83,"latest_published_at":88},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":90,"slug":91,"count":92,"latest_published_at":93},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]