The question behind AI panic: What happens to me if it works?

Consider the scenario where your supervisor commends a report you completed using AI. A task that traditionally took an afternoon is now accomplished within an hour, and you confirm that the output meets quality standards.

Should you disclose your use of AI?

If you choose to inform your supervisor, you could become the go-to expert for the new technology, possibly leading to better assignments, a pay increase, or even a promotion. On the other hand, keeping this information to yourself allows you to continue being recognized for your exceptional performance.

However, a more concerning possibility arises. Your supervisor may evaluate your productivity and ponder whether a task typically managed by a team of three could now be delegated to just one person. What initially felt like a significant achievement now invokes anxiety. And remaining silent does not eliminate risk: a colleague could independently find a way to automate part of your role.

This scenario encapsulates the current sense of panic. It's not a matter of whether people can effectively utilize the tool, but rather how demonstrating increased efficiency might lead to job reductions. The concern persists even among those proficient in AI, stemming from the question of what occurs once the technology is effectively implemented. This ambiguity cannot be resolved through mere explanations of the technology itself.

A survey conducted by Glean, a workplace AI company, involving 6,000 predominantly computer-based workers across the U.S., U.K., and Australia, revealed varying impacts of AI. While some reported enhanced productivity and improved work quality, a notable 36% admitted to concealing the extent of AI's benefits.

The anxiety of appearing replaceable was echoed in a 2024 survey conducted by Microsoft and LinkedIn, which found that over half of AI-utilizing workers expressed concerns that employing AI for critical tasks could foster that very impression.

Workers have faced this bargain before

This dilemma is not new; over a century ago, factory managers encountered similar issues. Workers would intentionally withhold their full capabilities to prevent management from realizing their potential output.

Frederick Winslow Taylor addressed this behavior in his 1911 treatise on scientific management. In his observations, some workers were compensated based on the number of items produced, creating a disincentive for increased productivity, as employers would subsequently lower the payment per item. As a result, workers faced the challenge of having to produce even more to maintain their previous earnings.

This phenomenon, known as "soldiering," was something Taylor sought to mitigate, though he recalled empathizing with workers who faced pressure to produce more.

Now, envision a manager requesting all employees to showcase their AI efficiencies. For management, it presents an opportunity for team learning; for employees, it may feel like an implicit prompt to illustrate how the firm could operate with fewer staff members. Ultimately, applause does little to alleviate such fears.

As the leader of a research center in Norway, I regularly employ AI to streamline tasks such as budget management and minute-taking for meetings. However, the time saved often corresponds to the time required for subsequent budget evaluations.

A colleague using the same technology to leave work promptly encounters a similar dilemma. When the reward for sharing such efficiencies results primarily in increased workload, the initial relief quickly dissipates.

Give workers a reason to share

On October 2, 2026, Norway's government, junto with eight employer and union organizations, formalized an AI agreement effective until 2029. This agreement mandates employee involvement during AI implementations while affirming existing rights and ensuring benefits extend to workers, businesses, and society alike.

This collaborative approach has historical precedence. In the early 1970s, computer scientist Kristen Nygaard partnered with Norway's Iron and Metal Workers' Union to introduce computers into workplaces, advocating for participatory design to engage users in decisions regarding system deployment.

While the new agreement serves as a declaration of intent, it does not guarantee wage increases or a shared portion of saved resources. The tangible impact of this agreement on workplace dynamics remains to be seen, particularly given the stark differences in union membership: in 2025, 52.6% of Norwegian workers were unionized, contrasting sharply with only 10% in the U.S. workforce and a mere 5.9% in the private sector.

While American employers can replicate the conversation, the key issue is whether employees have a decisive voice in determining what AI implementation means for their roles, rather than being consulted post-decision.

Envision an alternative approach where a company invites employees to demonstrate their AI efficiencies with a six-month trial. During this period, participants are assured they will not face any loss in pay or employment due to their contributions. Employees could influence how any productivity gains are utilized, fostering discussions with management regarding outcomes before targets are adjusted.

This initiative could empower workers to pursue paid training or allocate time to address backlog issues that have necessitated late hours, while the company might aspire to expand its business. The objective is to engage in these discussions before the benefits have already been designated elsewhere.

Such commitments should be documented, providing employees with a mechanism to contest unmet assurances. While such commitments would not necessarily constrain a company that later opts for a leaner team, they can afford employees a motivating factor to disclose their findings and innovations.

Taylor offered higher compensation to employees who achieved management's goals. Today, employers must also tackle the questions arising from those who uncover more efficient work methods.

That employee may already possess a compelling demonstration of their accomplishment. The decision to showcase it to leadership is often influenced by the experiences of previous employees who identified ways to streamline operations.