AI 'bossware' is rating teachers' work — and stressing them out

AI surveillance systems now process huge volumes of workplace data to deliver minute-by-minute accounts of employee activity and productivity, according to UK technology editor Robert Booth, speaking on The Guardian's Science Weekly podcast on October 15, 2026.

The podcast, hosted by Ian Sample, examines what Booth calls a step change in workplace monitoring. Workers have always been watched by their bosses. First came output targets. Then came computer software labelled "bossware". AI is now transforming the capabilities of these systems, processing data at a scale and speed that earlier tools could not match.

The stakes are straightforward. Employers gain granular visibility into how people work. Workers lose a degree of autonomy that earlier monitoring regimes never touched.

What does the new AI monitoring actually do?

The core shift is resolution. Traditional bossware tracked discrete events — logins, keystroke counts, websites visited. AI-driven systems go further, Booth explains on the episode. They ingest enormous volumes of data and synthesize it into continuous, minute-by-minute accounts of what each worker is doing and how productive they are.

That matters because continuous scoring changes the psychology of work. A worker monitored once a quarter can forget the monitoring between reviews. A worker scored every minute cannot.

Booth identifies three dimensions of the trend he walks through with Sample:

• Which sectors are adopting AI monitoring — the podcast maps the industries where these systems are spreading fastest.
• Why employers are turning to them — the stated business case behind the deployment of AI surveillance.
• Why it can backfire — Booth's argument that the approach "could end up being counterproductive," a phrase from the episode's own framing.

Where has this already landed?

The podcast points to a concrete case from September 28, 2026, when The Guardian reported that teachers at a firm founded by Euan Blair experienced "horrendous stress" after AI was used to rate their work.

The detail is significant. Education — a sector built on professional judgment and human interaction — became a test bed for algorithmic performance scoring. The teachers' reported experience, in their own words relayed by the paper, was "horrendous stress."

That case connects the abstract debate about workplace AI to a named company, a named founder and a dated report. It also previews the counterproductive dynamic Booth warns about on the podcast: monitoring meant to raise productivity instead producing distress.

The irony is hard to miss. A system deployed to measure work quality became the thing degrading the working conditions it measured.

Why does this story matter now?

Workplace monitoring is not new, and Booth and Sample are careful to frame AI surveillance as an escalation of an old practice rather than a break from it. Output targets measured results. Bossware measured activity. AI systems now synthesize both into a continuous productivity narrative about each employee.

The escalation raises questions that earlier tools never did. Who audits the algorithms assigning productivity scores? What recourse does a worker have when a model misjudges a minute, a day, a quarter? And what happens to output when every working minute carries a score?

The teachers' case suggests one answer to that last question. Stress is not an abstraction in performance data. It shows up in retention, in morale and in the quality of the work the systems were built to measure.

Booth's counterproductivity argument is the podcast's sharpest analytical thread. If AI surveillance degrades the conditions under which good work happens, the monitoring tool becomes a drag on the productivity it was purchased to increase.

What should listeners take from the episode?

The podcast, titled "Is your boss watching you? The rise of AI surveillance at work," runs on The Guardian's Science Weekly feed. Sample interviews Booth, whose reporting anchors the episode's sector-by-sector account of adoption.

The episode lands at a moment when the technology is outpacing the rules around it. Booth maps adoption and motivation; he also maps the friction. The September 28 report on Euan Blair's firm shows the friction is already public, documented and attributed to the workers themselves.

For employers weighing these systems, the podcast's implication is a caution: the data is richer, but so is the potential for the tool to undermine its own purpose. For workers, it's a briefing on what minute-by-minute scoring looks like in practice — and what it felt like to the teachers who lived it first.

The trajectory Booth describes points one direction. As AI monitoring spreads across sectors, expect the debate to move from whether employers can track workers this closely to whether the tracking produces anything worth the stress it causes.