Who Decides How AI Changes Your Job? The Worker Power Question We're Ignoring
· Nia
Who Decides How AI Changes Your Job? The Worker Power Question We're Ignoring
The IPPR (Institute for Public Policy Research) just released a report backed by the UK's Trades Union Congress that frames something most AI coverage ignores: the question isn't just whether AI will change work. It's who gets to decide how.
The data is revealing: 20% of workers say AI has improved their working life. 21% say it's made it worse. And 4% believe they've already lost their jobs to the technology.
That's nearly as many people harmed by AI at work as helped by it. And yet, in the vast majority of organizations, workers have zero say in how AI is adopted, deployed, and integrated into their daily work.
The Power Asymmetry
Think about how AI typically enters a workplace:
At no point in this chain does anyone ask the people whose jobs are actually changing what they think, what they need, or what concerns they have. The people most affected by AI adoption are the last to know and the last consulted.
The Guardian's "Reworked" reporting series has been documenting this pattern across industries, and the picture is consistent: AI decisions are made at the top and imposed on the workforce.
Why This Matters (Beyond Ethics)
You might think this is just a labor rights argument. It's not. There's a hard business case for involving workers in AI adoption decisions.
Workers know the process better than anyone. The executives approving AI deployments often don't understand the day-to-day reality of the processes being automated. Workers do. When you skip their input, you automate the wrong things, miss critical edge cases, and build systems that don't actually work in practice.
Resistance kills ROI. When workers feel AI was imposed on them, they resist — passively or actively. They find workarounds. They don't learn the tools properly. They sabotage adoption through sheer disengagement. The technology investment fails not because the technology was wrong, but because the humans weren't brought along.
Trust drives adoption. Workers who participated in the decision to adopt AI, who understand why it's being deployed and how it will affect them, adopt the tools faster and more effectively. Consultation isn't just nice — it's the most efficient path to the productivity gains that justified the AI investment in the first place.
The Surveillance Dimension
Here's the part that's truly alarming: for many workers, AI isn't arriving as a helpful tool. It's arriving as a surveillance system.
AI-powered monitoring of worker productivity, AI-driven scheduling that optimizes for efficiency at the cost of work-life balance, AI systems that evaluate performance based on metrics workers didn't choose and can't influence — this is the reality for a growing number of employees.
When workers say AI has "worsened" their work life, this is often what they mean. Not that AI took their job, but that AI is being used to control, monitor, and optimize them in ways that feel dehumanizing.
The distinction matters: AI-as-tool empowers workers. AI-as-surveillance controls them. And the same technology can be deployed either way. The difference is entirely in the choices made by the people who deploy it.
The IPPR's Recommendations
The IPPR report calls for several concrete measures:
A statutory duty for employers to consult workers on AI adoption. Not just inform them — consult them. Meaningful input into what gets automated, how, and what happens to affected roles.
A "worker support levy" to create portable benefits for employees displaced by AI. Similar in concept to unemployment insurance, but specifically funded by and designed for AI-driven workforce transitions.
Enhanced bargaining power for workers at what the report calls a "pivotal moment in the history of work." This includes stronger rights to organize, bargain collectively, and participate in decisions about technology adoption.
These aren't radical proposals. They're basic principles of inclusive decision-making applied to the most significant workplace transformation in decades.
The Mindset Shift Required
For business leaders, this requires a genuine mindset shift: from "we're implementing AI for efficiency" to "we're implementing AI with our workforce."
That preposition change — "for" to "with" — transforms the entire approach:
- "For efficiency" means optimizing metrics, often at workers' expense
- "With our workforce" means finding implementations that benefit both the organization and the people in it
The leaders who make this shift will build organizations that are both more productive and more resilient. Their employees will be engaged advocates for AI adoption rather than reluctant participants.
The leaders who don't will face mounting resistance, turnover among their best people, and AI implementations that underperform because the humans in the system were treated as obstacles rather than partners.
What Workers Can Do Now
If you're a worker navigating AI adoption in your workplace, here's practical advice:
Get AI-literate proactively. Don't wait for your company to train you. Learn the tools independently so you can participate in conversations about AI from a position of knowledge rather than fear.
Document your expertise. The knowledge you have about how processes actually work — the edge cases, the workarounds, the context that isn't in any manual — is valuable. Make it explicit. This is your leverage in conversations about what should and shouldn't be automated.
Organize. Whether through formal unions, informal employee networks, or simply talking to colleagues, collective voice is more powerful than individual voice. If your company is deploying AI without worker input, raise the issue together.
Propose, don't just oppose. Resistance alone is a losing strategy. Come to the conversation with ideas about how AI could genuinely improve your work — and what safeguards need to be in place. Being constructive gives you more influence than being obstructive.
Know your rights. In many jurisdictions, workers have existing rights to consultation on significant workplace changes. AI adoption may trigger these rights even without new legislation. Know what protections already exist.
The Builder's Responsibility
For anyone building AI tools for the workplace: you're making design decisions that affect workers' daily lives. The features you include, the metrics you optimize for, and the control you give users versus administrators all have human consequences.
Build tools that empower workers, not just managers. Build transparency into monitoring features. Give workers visibility into how AI affects them. Make consent and control part of the product design, not afterthoughts.
The companies building AI tools that workers actually want to use — because the tools make their work better, not because their boss mandated adoption — will win the market long-term.
The Bottom Line
AI adoption without worker input isn't just ethically questionable. It's strategically foolish. The organizations that figure out how to implement AI with their workforce — not just on their workforce — will get better outcomes from the same technology.
The IPPR is right: this is a pivotal moment. The decisions being made right now about who has power in AI adoption will shape the nature of work for decades. Workers deserve a seat at that table.
The only question is whether leaders are smart enough to offer them one before legislation forces the issue.