[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-human-help-makes-ai-agents-better-but-only-on-its-own-terms":10,"sections":40},{"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":30,"tags":31,"sources":35,"feedback":39,"feedback_at":22,"cost_usd":39,"total_tokens":39},3686,"human-help-makes-ai-agents-better-but-only-on-its-own-terms","Human Help Makes AI Agents Better-But Only on Its Own Terms","A new benchmark finds human input boosts LLM agent outcomes, but gains hinge on timing, channel, and who is doing the intervening.","A new benchmark called HAS-Bench puts a number on what it actually costs-and gains-when humans step into AI agent workflows.\n\nResearchers introduced HAS-Framework, a graph-based system that treats both humans and LLM-powered agents as first-class participants, each with defined roles, permissions, and communication paths. On top of that, HAS-Bench runs evaluations across six task domains, measuring not just whether the job got done but how well the collaboration worked-tracking clarification quality, feedback use, safety, initiative, and interaction overhead. The core finding: human participation can meaningfully improve task completion and help agents recover from failure, but only when the timing, channel, and persona are right.\n\nThat conditionality is the part worth paying attention to. Most AI agent benchmarks treat human involvement as either fully present or fully absent. HAS-Bench instead models it as a dial with a lot of positions in between, which is much closer to how these systems actually get deployed in workplaces. The results suggest that poorly configured human participation can add cost without adding value-a finding that cuts against the reflexive assumption that more human oversight is always better.\n\nThe AI agent space is crowded with benchmarks right now, but most measure what an agent can do alone. HAS-Bench is an attempt to measure the harder question: what happens when the agent is not alone.","[\"ai\",\"benchmarks\",\"ai-agents\",\"human-ai collaboration\"]","2026-07-07T04:00:00.000Z","2026-07-07T05:46:37.260Z","2026-07-07T05:46:40.075Z","published",null,[24],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"The headline and dek are vague and placeholder-quality — 'how well they actually collaborate' restates the obvious without naming the specific metrics or finding; rewrite both to lead with the concrete result (human input improves outcomes but only under specific conditions of timing, channel, and persona) and sharpen the dek to match.","resolved","ai",[30,32,33,34],"benchmarks","ai-agents","human-ai collaboration",[36],{"name":37,"url":38},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.04329",0,{"sections":41},[42,46,51,56,61,66,71,76,81,85,90,94,99,104],{"name":43,"slug":30,"count":44,"latest_published_at":45},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":47,"slug":48,"count":49,"latest_published_at":50},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":52,"slug":53,"count":54,"latest_published_at":55},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":57,"slug":58,"count":59,"latest_published_at":60},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":62,"slug":63,"count":64,"latest_published_at":65},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":70},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":72,"slug":73,"count":74,"latest_published_at":75},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":77,"slug":78,"count":79,"latest_published_at":80},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":18},"Dev Tools","dev-tools",59,{"name":86,"slug":87,"count":88,"latest_published_at":89},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":91,"slug":92,"count":88,"latest_published_at":93},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":95,"slug":96,"count":97,"latest_published_at":98},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":100,"slug":101,"count":102,"latest_published_at":103},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":105,"slug":106,"count":107,"latest_published_at":108},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]