Clarke & Joffe (2025) found creative-agency workers must continuously configure and repair AI role boundaries — a related but distinct phenomenon from release work

generative-ai-at-workcscwaccountability-labourboundary-workcreative-workcorroboration

In "Beyond Replacement or Augmentation" (arXiv:2505.18938, submitted to CSCW 2025, acceptance unconfirmed), Michael Clarke and Michael Joffe studied how international creative-agency workers reconfigure the division of labour with generative AI, finding that "workers must continuously configure and repair AI role boundaries to maintain workplace intelligibility and accountability." The accountability-preserving labour they document is continuous, discursive, and collaborative — a team's ongoing renegotiation of what role the AI plays within shared work.

This is independent of the Abdulhamid et al. branch by the single-source cap's test — a different author group, a different venue, a different population (creative-agency workers rather than Kenyan/Nigerian SMB workers). It documents a genuinely related phenomenon: workers doing extra accountability labour around generative-AI outputs. But it is not confirmation of release work as specifically defined. Clarke & Joffe's "boundary repair" is ongoing and cooperative; release work is the one-time, often solitary certification of a single artifact's defensibility at the moment it crosses to an external audience — and Abdulhamid et al. are explicit that release work "can be discharged by one worker with no cooperative activity to articulate," a claim the collaborative framing here does not test. The two studies are compatible but not confirmatory of each other's taxonomy: the vault records them as independent CSCW observations that accountability labour around AI outputs recurs, not as two witnesses to the same construct. See question-corroborate-abdulhamid-2026-generative-ai-workable-at-work-single-source.