CITATION — REFERENCE ENTRY

AI Agents Push Humans Out of the Loop — arXiv

Revision 65aa1872-a62c-4304-964f-7d45c343e8bf · 9/25/2026, 5:46:34 PM UTC
Key
mitchell-ghosh-passi-2026-ai-agents-out-of-loop
Authors
Mitchell, Margaret; Ghosh, Avijit; Passi, Samir
Issued
2026-8-24
Type
article
Container
arXiv
Raw CSL JSON
{
  "DOI": "10.48550/arXiv.2608.23642",
  "URL": "https://arxiv.org/abs/2608.23642",
  "type": "article",
  "genre": "Position paper preprint",
  "title": "AI Agents Push Humans Out of the Loop",
  "author": [
    {
      "given": "Margaret",
      "family": "Mitchell"
    },
    {
      "given": "Avijit",
      "family": "Ghosh"
    },
    {
      "given": "Samir",
      "family": "Passi"
    }
  ],
  "issued": {
    "date-parts": [
      [
        2026,
        8,
        24
      ]
    ]
  },
  "number": "2608.23642",
  "container-title": "arXiv"
}

Claims

  1. Mitchell, Ghosh, and Passi list cognitive surrender alongside cognitive dependence and cognitive debt as recent terms for skill degradation from AI use, and propose rotating overseers between AI-assisted and unassisted tasks to prevent cognitive surrender from prolonged exposure to agentic AI.
    "Recent terminology has crystallized additional patterns: "cognitive dependence" [44], "cognitive debt" [65], and "cognitive surrender" [102]. [...] Rotations move users between tasks, and between AI-assisted and unassisted versions of the same task, to prevent both fatigue and the cognitive surrender from prolonged exposure to agentic AI."
    Quote language: en
Available in