AI can change what a person feels known for at work. Producing a strong first draft may become easier, while the expertise behind choosing, correcting and defending it becomes harder for others to see. Some people experience that as relief or a new opportunity. Others question the value of skills they spent years developing.
For a manager, the useful response is to make the person's contribution visible and give confidence a basis in work they can explain. Avoid assuming that enthusiasm means competence, or that discomfort means resistance. The broader people and change guide helps connect support to the employee's actual experience.
Recognize the contribution that has become harder to see
In an August 2026 Reddit discussion about creative work, a freelancer with more than 20 years of self-reported experience described clients bringing extensive AI-generated briefs as though the professional work were already done. The freelancer still saw problems with tone and quality, but felt less confident about the value clients placed on their contribution. They used AI themselves where it helped; the concern was not simply dislike of the tool.
Another Reddit user in the discussion described receiving AI text from executives and knowing where to push back. The original poster clarified that their current clients were mostly entrepreneurs, unlike their former corporate setting. The ability to challenge a draft depends partly on the working relationship. These accounts should not be collapsed into one story about all employees losing confidence.
Recognition can also become uncomfortable when the output is good. In an August 2026 firsthand account from Perceptyx, Sarah Jorgenson described receiving praise for a presentation analogy that ChatGPT had supplied. The useful result coexisted with questions about pride and her own contribution. The account comes from a commercial employee-listening vendor, rather than an independent study of how often this happens.
There is a different reaction in Francesca Leader's August 2026 X post and replies. She described feeling more legitimate as an artist in contrast to AI-generated work, while retaining her own writing practice. That is a personal reflection, not evidence of a clinical change. It challenges the assumption that AI necessarily diminishes everyone's sense of professional value.
Ask what the person feels they have gained or lost. It may be satisfaction in making something, recognition from a client, independent problem-solving or confidence that their judgment still matters. A general reassurance about the future of work will not identify which of those needs attention.
Separate confidence in AI from confidence in your judgment
A 2024 study of one US software organization examined 35 survey responses and 11 interviews gathered in 2023. Junior developers valued quicker answers but also worried about independent learning and ownership. Senior developers emphasized control and accountability, with mixed views on where the tools helped. Both groups had reasons to value assistance and protect their contribution.
The small, early study does not establish a rule for every junior or experienced employee. It does suggest different questions to ask. A junior may need practice reasoning through a problem. An experienced professional may need space to challenge generated recommendations and explain the context they missed.
Confidence in the tool is a separate issue. A CHI 2025 survey of 319 knowledge workers, including Microsoft researchers, analyzed 936 reported examples of generative AI use. Greater confidence in AI for a task was associated with less reported critical-thinking effort; greater confidence in one's own ability was associated with more. These were self-reports, not a test showing that AI caused skills to decline. Subjective confidence also does not establish objective expertise.
For a team, a practical next step is to look at the work behind the confidence. Ask the author to explain an evidence choice, a rejected recommendation and a remaining uncertainty. Neither study tests whether this review approach improves confidence. Keep the discussion proportionate to the consequences of the task.
Give judgment a concrete place in the work
Suppose Jamie Lee, an experienced designer at a retailer, reviews a product page with Alex Morgan, a junior colleague. Their inputs include the existing page, approved product information and customer-support reports about shoppers misunderstanding delivery eligibility. They use an approved assistant to suggest page revisions from a permitted sample.
The generated advice proposes a cleaner layout but removes a delivery condition shoppers need before ordering. A polished page would still create the same customer problem. Jamie explains why the condition matters; Alex revises the proposal and checks it against the approved information.
The finished work should make three contributions explainable:
- Choosing the problem. Alex can explain why delivery eligibility, rather than visual neatness alone, is the priority.
- Rejecting the unsuitable suggestion. The team can identify the condition the assistant removed and the source that requires it.
- Checking the revision. Alex compares the final wording and placement with the approved delivery information before handing it to the page owner.
Jamie still contributes professional judgment, and Alex gets a chance to practise it. Generating another version without discussing the rejected advice would miss that opportunity. The review need not become a performance in which the junior must prove they never needed help. Its purpose is to build the ability to explain work someone else will depend on.
On the LinkedIn article about performance engineers' changing work, Sean Varnham described trying to instil a user-experience and business focus in his team, adding that it was taking time. The self-described principal performance engineer's comment is a useful example of professional development beyond producing technical output. It does not establish that a role is safe from replacement or that every team has made the transition.
Change the review conversation and the recognition
A manager can acknowledge a person's contribution without promising a promotion, preserving every task or insisting they become enthusiastic. Be specific about what the work requires and which decisions belong to the professional.
| What you observe | What to investigate | A useful response |
|---|---|---|
A junior submits a fluent draft but cannot explain a key choice. | Which reasoning step was skipped, and what source should guide it? | Revisit that step together and let the junior make and check the next decision. |
An experienced employee says their contribution is being overlooked. | Which judgment, relationship or correction has become invisible? | Include that contribution in feedback and explain why it matters to the finished work. |
Someone is highly confident because the AI answer sounds convincing. | Can they find the supporting evidence and identify a condition that would change the answer? | Review the evidence before relying on the output; arrange support where checking is difficult. |
Recognition should include the work that makes an output dependable: identifying a missing condition, choosing relevant evidence, correcting an unsuitable recommendation and explaining a decision to a colleague. The guide to performance systems that support useful AI adoption explores how to recognize judgment and quality alongside use.
Keep actual employment concerns answerable too. If the question is whether a role or team will change, use the separate guide to discussing job uncertainty honestly. A conversation about professional pride cannot substitute for a clear answer about a staffing decision.
Start with one recent output. Ask its author what they contributed, what the assistant contributed and which check made the result acceptable. Use the answer to choose a concrete change in practice or feedback. Do not infer the state of someone's professional identity from a tool-use count.
Questions and answers
Does AI always weaken professional identity?
AI does not produce one inevitable change in professional identity. Someone may value being able to do more, feel that a hard-earned skill is less recognized, or place renewed value on their own craft. Individual accounts and small-context studies show mixed experiences; they do not establish how every profession will respond.
Ask which part of the work has changed and what the person feels known for now. A client's dismissal of design judgment calls for a different response from a junior's lack of practice. Use the contrasting experiences to frame a specific conversation.
How can employees keep confidence when AI contributes to their work?
Employees can ground confidence in decisions they can explain and work they can check. Identify the problem being solved, the evidence used, a generated suggestion that needed changing and the final acceptance check. Help from AI does not by itself remove the person's contribution, but a confident tone is not proof that the output is dependable.
For a retailer's product page, that means explaining why delivery eligibility matters and checking that the revised wording preserves the approved condition. Practise the three contributions in the page review with a colleague when needed.
Do junior and experienced employees need different support?
Junior and experienced employees may need different support with AI, but seniority alone cannot tell a manager what is missing. A junior may need practice choosing evidence and explaining a recommendation. An experienced professional may need their contextual judgment to be heard, or help evaluating a tool in unfamiliar work. Either can benefit from assistance or be overconfident in an answer.
Look at a recent task before assigning training. Ask the person to explain a key choice and where checking became difficult, then arrange support around that step. The research on ownership and confidence explains why these questions should remain separate.
What should a manager do when an employee feels less valuable?
A manager should listen for the specific contribution the employee feels has lost value, then examine how it is recognized in real work. Ask about a recent output, who relied on it and what judgment or correction the employee supplied. Change feedback or the review arrangement where that contribution has become invisible, rather than assuming the person needs encouragement to use AI more.
Avoid promises about job security that you cannot support. Deal with actual staffing questions directly, and use the review-conversation choices to agree one practical next step for the work.
