The AI‑based performance system introduced at Multiverse
Euan Blair’s technology‑training firm, Multiverse, has deployed an AI model that analyses classroom transcripts and assigns each instructor a “risk status” and a “percentage confidence score”. The algorithm flags a range of behaviours, from delayed reconnection after a technical glitch to the use of filler phrases such as “sort of” and “kind of”. Teachers describe the mechanism as “remorseless” and “unnerving”, noting that the constant scrutiny has led to sleep loss and a need for counselling.
Multiverse, co‑founded by Blair, supplies AI and digital‑skills training to the NHS, local authorities, universities and private firms, drawing most of its funding from the government’s apprenticeship levy. The company’s valuation stands at £1.6 bn, although its latest accounts show a turnover of £118 m and losses of £28 m. Blair, who retains a large minority stake, has a personal fortune estimated at £350 m.
The AI does not replace human managers; instead, it is intended to “direct human time to where it’s needed most”. According to a spokesperson, the system “improves the experience of learners and customers” and only human reviewers ultimately write performance assessments. However, the firm has not directly addressed the staff’s complaints about anxiety and the perceived intrusiveness of the tool.
Staff wellbeing and the toll on instruction
One instructor, speaking on condition of anonymity, said the recent changes have produced “absolutely horrendous” stress levels. “I haven’t ever been in a position where I’m waking up in the middle of the night,” they added. The shift from monthly human observations to near‑continuous AI scrutiny has altered teachers’ focus. “Your attention is no longer delivering [the lesson] and listening to the learner [but] ‘am I saying the right thing so the AI doesn’t mark me down for something stupid that I’ve mistakenly said’,” another teacher explained.
The AI’s criteria extend beyond technical glitches. It penalises vague answers, prolonged one‑to‑one interactions that sideline the group, and even visible confusion when an instructor self‑corrects. “If the instructor is visibly confused themselves or had to self‑correct while giving the instruction, that confusion itself counts as evidence the instruction was not clear the first time,” the system’s guidelines state. Teachers report that managers now question them for suspected “digressions” that were in fact responses to student queries, creating a climate of fear and self‑censorship.
Industry reaction and calls for regulation
The Communication Workers Union’s national tech officer, John Chadfield, warned that “the public would be horrified to hear of bosses deploying deeply intrusive surveillance on workers”. He argued that the tech sector is “in serious need of further regulation” to ensure AI is used responsibly, not to “shut off job opportunities for young people, degrade the craft of skilled workers and make astronomical profit for a few”.
Earlier examples of AI monitoring in other sectors have drawn similar criticism. Burger King began tracking restaurant staff’s customer interactions this year, while Meta paused keystroke monitoring amid privacy concerns. Research published this week suggests AI surveillance can backfire, leading to “cognitive surrender” and reduced job satisfaction among UK tech workers.
Multiverse’s defence and the broader AI narrative
Multiverse maintains that the AI tool is a quality‑assurance aid, not a replacement for human judgment. The company says instructors will eventually gain dashboard access to the data collected, and that all performance reviews remain the responsibility of human managers. Blair, alongside his father Tony, has been a vocal advocate for expanding AI adoption across the UK economy, arguing that “every single worker is going to need to be retrained” and that firms should be “as AI‑pillled as possible”.
The firm’s AI monitoring agent, which incorporates models from Anthropic and OpenAI, is programmed to detect patterns of filler words and hedging, emphasising that isolated instances are not penalised unless they become “pervasive repeated” across a session. Multiverse’s spokesperson described the system as providing “improved advice on development and feedback – exactly how any learning organisation should be deploying AI”.
Why it Matters
The controversy surrounding Multiverse’s AI monitoring highlights a growing tension between technological efficiency and employee autonomy. As organisations increasingly turn to algorithmic oversight to optimise performance, the human cost—manifested in heightened stress, eroded morale, and a chilling effect on pedagogical creativity—cannot be ignored. This case underscores the urgent need for robust regulatory frameworks that balance innovation with the protection of workers’ mental health and professional dignity, ensuring that AI serves as a tool for empowerment rather than a mechanism of relentless surveillance.
