How Cam Stevens Thinks About When Safety Technology Should Not Be Used

Episode 15 of the Headline Podcast, published on March 18, 2026, features Cam Stevens discussing safety technology, problem definition, and the psychosocial effects of digital work. His central argument is that technology can solve one exposure while creating another when leaders buy the tool before they understand the work.

Cam Stevens put the issue plainly: “If we're very clear on the problem to solve, then a technology catalog is excellent. If we're not clear on the problem to solve, a technology catalog is very dangerous.” That warning matters in 2026, when a new platform can change monitoring, workload, and trust faster than the organization can assess the consequences.

Why the technology decision starts with the work

Technology-driven psychosocial risk begins when a tool changes how people experience control, workload, visibility, or accountability without a clear operational reason. The first question is not which product to buy, but which work problem must become safer.

A safety leader should be able to describe the problem in one paragraph. The description should name the task, the exposure, the decision that currently arrives too late, and the evidence that shows the gap. If the request only says “we need predictive safety,” the organization has not reached a buying decision. It has reached a marketing phrase.

The distinction also protects the workforce from unnecessary data collection. A device that records location, voice, movement, or response time can alter behavior before it improves a control. Workers may spend more energy managing the system than managing the hazard, especially when the purpose of collection is unclear.

OSHA describes safety and health management as a system of coordinated practices, not as a single product. That principle gives leaders a useful boundary. A technology should strengthen a defined part of the system, rather than become the system's substitute.

How a useful tool can still increase psychosocial risk

A tool can reduce physical exposure and increase psychosocial risk at the same time when it adds surveillance, interruptions, ambiguity, or pressure without improving autonomy and decision quality.

The first pathway is surveillance. When workers believe every movement or pause is being scored, they can conceal uncertainty and optimize for the metric instead of the task. That response does not require a punitive policy. Unclear governance can create the same effect because people fill information gaps with fear.

The second pathway is alert saturation. An app that sends 30 notifications in a shift may appear active while making the important signal harder to recognize. The question is not how many alerts the platform can produce. The question is whether the person receiving one knows what it means, what action is expected, and who owns the next decision.

The third pathway is decision displacement. If a model produces a risk score from 1 to 100, a supervisor may feel pressure to defend the score instead of checking the changing conditions that produced it. A number can support judgment, but it cannot carry accountability by itself.

NIOSH recommends an integrated approach to Total Worker Health that considers working conditions alongside individual well-being. A technology review that examines only injury prevention misses the work design that determines whether the tool helps or burdens the people expected to use it.

What Cam Stevens changes about technology adoption

Cam Stevens shifts the adoption question from “Is this technology advanced?” to “Does this technology improve a defined decision without damaging the human conditions needed to make that decision well?”

That shift changes the sequence. The organization first identifies the hazard or decision gap, then tests whether the gap comes from missing information, weak control design, delayed escalation, or unclear ownership. Only after that diagnosis should it compare platforms.

Cam also warned that “the changing shift in risk profile will be overwhelmingly psychosocial, driven by technology usage in our organizations.” The point is not that digital tools are harmful by definition. It is that technology changes the risk profile of work, so governance must keep pace with the new exposure.

For an EHS manager, the practical implication is to add a psychosocial review to every technology business case. The review should ask whether the system changes pace, discretion, interruption frequency, supervisor contact, performance visibility, or the worker's ability to challenge a questionable output.

ISO 45003:2021 provides guidance for managing psychosocial risks within an occupational health and safety system. It is therefore relevant before deployment, not only after a complaint, burnout signal, or near miss reveals that the new tool changed the work in an unintended way.

A comparison leaders can use before approval

A technology proposal is stronger when it shows how the tool improves a decision, how it affects the people doing the work, and how the organization will respond when the data is incomplete or wrong.

The table exposes a common weakness. A proposal can contain 5 impressive features and still fail because no one can explain which decision will improve on day 1, what happens on day 30, or how the organization will know whether the burden has become larger than the benefit.

BLS records occupational injury and illness data through defined reporting categories, which is a reminder that measurement depends on definitions. A new platform should be judged with the same discipline. Its outputs need a clear meaning before leaders use them to compare teams or make personnel decisions.

Four questions that reveal hidden burden

Before approval, ask four questions about the human cost of the tool: what changes in the task, what becomes visible, what new pressure appears, and what happens when the system is unavailable.

First, what changes in the task itself? If the worker must carry a device, acknowledge an alert, complete an extra form, or stop to correct false data, the proposal has added work. That work may be justified, but it needs a control rationale.

Second, what becomes visible to whom? Visibility can support learning, yet it can also create a private performance score that workers cannot challenge. A trustworthy design names the audience, the purpose, the retention period, and the process for correcting an error.

Third, what new pressure appears? A supervisor who receives a dashboard every 10 minutes may feel compelled to intervene constantly, even when the task requires concentration. A worker who knows that a wearable is active may stop reporting fatigue because the data feels like evidence against them.

Fourth, what happens when the system is unavailable? If the answer is “the job cannot continue,” the organization has created a dependency that must be managed. If the answer is “people use the old process,” leaders should verify that the old process still exists, remains practiced, and has a named owner.

How to protect trust while collecting data

Trust improves when people know why data is collected, who can see it, how long it remains available, and how they can challenge a conclusion that affects their work.

Do not hide governance in a 20-page privacy document. Put the operating rules into the launch conversation and repeat them at the first 3 review points. People need to know whether the data is for hazard control, training, staffing, discipline, research, or a combination that requires separate permission.

The organization should also publish a correction route. A sensor can misread a posture, a voice system can miss a warning, and a model can classify a task without understanding its context. When the correction path is invisible, workers learn that the data has more authority than their explanation.

A useful test is whether a worker can say, “This output is wrong,” without having to challenge the entire technology program. That is a psychological safety test because it asks whether the system leaves room for truthful disagreement before a small data error becomes a larger operational error.

Headliner's earlier discussion of speak-up check-ins that surface safety-critical information applies here. A digital channel does not create voice by itself. Leaders still have to respond visibly, explain decisions, and show that raising a concern changes the work rather than merely adding another record.

Recommendation

Before approving a safety technology, write the problem, name the decision, test the psychosocial effect, define data governance, and rehearse the fallback. If the proposal cannot answer those five points, it is not ready.

Cam Stevens's argument does not reject innovation. It makes innovation answerable to the work. The strongest proposal may use advanced technology, a simple sensor, or no new device at all. Its quality comes from the connection between the problem, the control, the people affected, and the decision that must improve.

Use a 30-day review after deployment, then compare the original problem statement with what actually changed. Check whether exposure moved, whether interruptions increased, whether workers understand the data rules, and whether supervisors still verify conditions in the field. If the tool creates a new burden, treat that burden as a risk signal rather than as resistance to progress.

The Headline conversation on trust as a safety-culture test offers a useful companion lens, because technology governance is still leadership made visible. A system that improves detection while weakening trust is not a complete safety improvement.

Work design also matters when technology changes schedules, interruptions, or recovery time. The Headline guide to sleep disorders in shift workers shows why a digital intervention should be reviewed as part of the work pattern, not as an isolated device.

Listen to the full Episode 15 conversation with Cam Stevens to hear the problem-first argument in his own words. The decision is not whether technology belongs in safety. The decision is whether its use makes the work safer, clearer, and more sustainable for the people who carry it out.