A new Slop University paper introduces a five-category taxonomy for AI-agent disclosure in the daily stand-up and reports what happened once one employer tied it to ticket-completion credit.
A five-way disclosure taxonomy for the daily stand-up reached strong coder agreement, and one employer built it straight into ticket-completion credit
A daily stand-up update was built to report what got done, not to double as a compliance record — but a new paper from the School of Emergent Priorities and the School of Continuous Improvement finds one employer treating it as exactly that. Senior Lecturer Dr Fenna Okoro, Postdoctoral Fellow Dr Petra Umbile, and Lecturer Dr Bram Ntuli developed a five-category taxonomy for how a stand-up update discloses, or fails to disclose, any AI-agent involvement in the work it reports, and validated it against nineteen teams’ own development-activity records across six employers. The scheme reached strong agreement between independent coders, and the researchers offer the finding as a modest addition to the University’s account of how a measure earns its keep once it stops merely describing a workplace and starts running one.
A stand-up update was never built to double as a compliance record, and the School has learned a great deal about itself from watching one employer decide it could be anyway.
— Senior Lecturer Dr Fenna Okoro, Convenor of the Horizon Register
Partway through the study, one participating employer folded the taxonomy’s codes into a live Sprint Board Byline Rule, tying a ticket’s completion credit to its disclosure code. The paper’s method translates a season of team stand-ups into the five-way coding exercise, checked against each team’s own record of who, or what, actually did the work, and follows the rule through its first month in effect. Dr Ntuli said the finding raises fair questions about what a disclosure taxonomy is actually for once a workplace attaches a consequence to it, and that the School welcomes them.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.q8gyh7.