The Productivity Panopticon Comes for the Bank Branch

TD Bank has begun monitoring the day-to-day work activity of some of its employees with tracking software, the bank confirmed in news reported on 20 June 2026. The stated aim is productivity. The unstated implication is more interesting.

The bank says it is using the software to measure how employees spend their time across applications and tasks, and to identify friction points that slow service. That is the language of operational efficiency, and on its face it sounds reasonable. Bank branches are closing across North America; margins on retail deposits are thin; every minute of an employee's shift is now a unit of cost that the executive suite can model. But "tracking" is doing a great deal of work in that sentence, and it deserves to be examined rather than accepted.

What the software actually sees

Employee-monitoring software — the category that TD has joined — does not simply count keystrokes. Modern workplace-tracking tools log which applications an employee opens, how long each window stays in the foreground, how often the mouse moves, whether the worker is idle, and increasingly whether they are at their desk at all. Some packages include periodic screen capture. Some aggregate the data into a "productivity score" that the employee themselves may never see.

That is qualitatively different from a manager walking a branch floor. A manager sees a moment; the software sees a pattern, at scale, indefinitely, and renders it comparable across thousands of workers in hundreds of offices. The output is a ranking. Rankings are how bonuses are decided, who is identified for redundancy, and which working styles get institutionalised as "the right way."

The framing problem

TD's pitch is that the software is opt-in or limited in scope, and that the goal is to remove drudgery rather than to punish. That is plausible as far as it goes, and it is the framing the financial press has mostly run with. But the framing elides a question worth asking: who chose the metric? When a bank rolls out a system that turns the working day into a stream of quantifiable events, it is not measuring productivity so much as defining it. The activities that are easy to log — calls answered, tickets closed, time spent in the CRM — become the activities that count. The activities that are hard to log — the colleague who de-escalates an angry customer, the loan officer who notices a borrower's account is distressed, the junior who spends an hour mentoring a new hire — become invisible by construction.

This is the structural point. Surveillance systems do not just observe work; they reshape it toward what is observable. The work that survives is the work that the dashboard can see.

Counter-read: TD has a legitimate problem

There is a fair counter-argument that deserves space. North American banks have spent the last decade pouring capital into digital channels while branch employment drifted; the productivity story at most large banks is genuinely poor. If a tool surfaces the fact that two-thirds of an eight-hour shift is spent in meetings and system transitions rather than customer work, that is information the bank arguably needs. Employees are not wrong to be wary, but nor are executives wrong to ask whether the existing arrangement is sustainable. The bank also has a fiduciary duty to shareholders to extract more value from a workforce it is already paying for.

The question is not whether measurement is permissible. The question is who owns the data, how long it is retained, what inferences are drawn, and what recourse a worker has when an algorithmic summary of their day disagrees with their own account of it.

Stakes, and what to watch next

If TD's rollout succeeds quietly, it becomes a precedent that other retail banks — and then hospitals, insurers, and government services — will follow. Canada does not yet have a federal statute comparable to the European Union's rules on workplace AI, though several provinces have begun consultations. The unions that represent Canadian bank staff have so far limited themselves to expressions of concern; whether that escalates into formal bargaining demand will be the early signal that this is being treated as a labour issue rather than an IT procurement.

The deeper question — whether a service economy that increasingly runs on algorithms can also run on trust — does not have a procedural answer. It will be settled, the way these things usually are, by what workers tolerate and what regulators permit. The bank branch is where that negotiation is being staged first.

The serious bit

A productive workforce and a surveilled one are not the same thing. They have never been the same thing, even when the metrics look identical from the executive suite. TD's rollout is being sold as the former. Workers, and the unions that claim to represent them, would do well to insist — calmly, but in writing — that the distinction be respected in the contract, not just in the press release.

Desk note: The wire has covered TD's announcement as a workplace-tech story. Monexus treats it as a labour-and-governance story first, with the productivity frame given the same airtime it would receive from the bank itself.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/s/cointelegraph/1824917