AI supporting well-being or AI surveillance: What psychology tells us about the workplace divide

There's a divide growing in how people experience AI at work. And it has almost nothing to do with technology.

Reading about Meta's mouse tracking software implementation this week and Nazrul Islam's piece in The Guardian about AI and worker surveillance really stopped me in my tracks and made me reflect on what is happening in workplaces right now.

Some people are using AI to clear the repetitive work so they can focus on thinking, relationships, and judgement. Others are being scheduled or monitored by it. Same technology. Completely different human experience.

Self-Determination Theory (SDT) (Deci & Ryan) says that motivation and wellbeing depend on three things: autonomy, competence, and relatedness. When AI supports those needs or provides the space so you can lean into those needs, people embrace it. When it undermines them, people disengage or leave.

In a workplace where AI is being used as a surveillance tool you don't have autonomy, you are afraid that you are going to be replaced by a bot, and you lose connection to your organisation and purpose.

In Meta's case, employees didn't just disengage. They revolted.

What happened at Meta

Meta installed mouse tracking software on employee computers to capture how people navigate screens, click through menus, and interact with applications. The company's stated reason was that they need real examples of how humans use computers to train their AI agents.

The problem? This happened in the same breath as announcing plans to cut roughly 20% of their workforce. Employees were being asked to generate the training data that would power the automation of their own roles, with no meaningful right to opt out.

Staff distributed anonymous flyers in meeting rooms and balanced them on toilet paper dispensers. The pamphlets asked a pointed question: "Don't want to work at the Employee Data Extraction Factory?" In the UK, Meta employees launched a formal union drive.

This isn't a story about flyers and protests. It's a story about what happens when all three pillars of human motivation collapse at once.

Autonomy: "I have no say in this"

SDT identifies autonomy as the need to feel in control of your own actions and choices. It doesn't mean working without rules. It means having a voice in how your work gets done.

At Meta, employees had no choice about the mouse tracking. No opt out. No consultation about how their data would be used. The International Labour Organization's 2025 research found exactly this pattern: AI deployment often intensifies work, increases surveillance, and limits worker autonomy. When that happens, motivation drops and psychosocial risk increases.

Now contrast that with a small business owner who says to their team member, "I noticed you spend a lot of time on those weekly reports. Want to try an AI tool and see if it helps?". That's the same technology. The person gets to choose. They get to experiment. They get to decide if it works for them. Their autonomy stays in tact!

Competence: "I'm being replaced, not developed"

The second pillar of SDT is competence, the need to feel effective and capable at what you do. Harvard Business Review's 2026 research confirmed that when AI supports a worker's sense of competence, engagement goes up. When it threatens it, people shut down.

At Meta, the mouse tracking software didn't make anyone feel more capable. It made them feel like their skills were being extracted. The message wasn't "we're investing in you." It was "we're capturing what you know so we don't need you anymore."

This is the critical difference between AI that builds capability and AI that harvests it. When someone on your team uses AI to produce a first draft of a client report in an hour instead of six and then adds their expertise and judgement to make it excellent, that's competence being enhanced. They're getting better at their job, not being sidelined from it.

Relatedness: "I don't feel like this organisation is on my side"

The third pillar is relatedness. The need to feel connected to the people and purpose around you. It's the reason people go the extra mile. Not because they're monitored, but because they care.

The HRD Australia article put it well in their closing line. Meta's employees didn't reach for flyers and federal labour law because the mouse tracking software was technically impermissible. They reached for them because no one had made them feel that the organisation was on their side. That's a people management failure.

When AI surveillance breaks relatedness, you don't just lose motivation. You lose loyalty, discretionary effort, and the culture that holds a team together. Research published in Scientific Reports in 2025 found a direct link between AI exposure and worker wellbeing, and the difference wasn't whether AI was present. It was whether workers felt their organisation supported them in using it.

Why this matters for small and medium businesses

You might be reading this and thinking you cannot relate to an organisation like Meta. You aren’t as big as them. You don't have mouse tracking software. And you're right. But the principle applies at every scale.

Every time AI is introduced into a workplace, it sends a signal. Either "we're giving you better tools to do your job" or "we're watching you more closely." Your team reads that signal instantly, even if you never intended to send it.

The good news? Small and medium businesses have a natural advantage here that large organisations struggle to replicate.

You know your people. You can have the conversation before you introduce the tool. You can ask someone what's eating their time and let them choose how to solve it. You can build trust in a week that would take a large organisation six months of workshops and external consultants to attempt to achieve.

When Deci and Ryan developed Self-Determination Theory, they found that environments supporting autonomy, competence, and relatedness don't just create happier workers. They create more motivated, more productive, more innovative ones. That's not a soft benefit. That's a competitive edge.

And it's one that smaller businesses are better positioned to create.

? The question for every business owner

Before you introduce any AI tool into your business, ask yourself three questions:

• Does this give my team more autonomy, or less?

• Will my people feel more capable because of this, or more replaceable?

• Does this strengthen our connection as a team or to our customers, or does it weaken it?

If the answer is yes, you're on the right track.

If the answer is anything else, you may want to rethink before you end up with your own version of anonymous flyers in the meeting room. Maybe not literally. But the disengagement, the quiet quitting, the loss of trust? That's the same thing.

The technology is the same. The experience is what you make of it.

---

Sources:

Matthew Sellers, "Meta workers storm their own offices over AI surveillance," Human Resources Director Australia, 13 May 2026

Nazrul Islam, "Forget the AI job apocalypse. AI's real threat is worker control and surveillance," The Guardian, 11 May 2026

International Labour Organization, "AI-driven intrusive surveillance and loss of autonomy at work linked to psychosocial risks for employees," 2025

Harvard Business Review, "Why Gen AI Feels So Threatening to Workers," March 2026

Scientific Reports (Nature), "Artificial intelligence and the wellbeing of workers," 2025

Deci, E.L. & Ryan, R.M., Self-Determination Theory (autonomy, competence, and relatedness as foundations of human motivation)

---

Tracy Attanasio is the Director of Nuovo Insights, helping small and medium businesses in South West Sydney navigate AI with a people first approach. She holds double degrees in Psychology.