Conversations With Your Employer

Demonstrating AI competence at work has a trap: if you show that AI can handle your tasks without making clear what you do with the freed time, you may accelerate the conversation about whether your role is still needed. The 2024 MIT Sloan and BCG Workforce Surveys repeatedly found that workers who hide their AI use perform worse over time than those who make it visible -- but visibility done badly is its own risk. This module covers how to frame AI use productively, what to avoid saying, and how to propose the augmented version of your role before your employer designs one for you.

The framing that works

The useful frame is not "I used AI to do X faster" but "I used AI to handle X, which freed me to do Y -- and here is the business result of Y." Y should be a higher-judgement activity: a decision you made, a relationship you developed, a problem you identified, an insight you surfaced. The argument is not speed, it is reallocation toward value. Speed alone invites the response "good, do it for two roles." Reallocation toward value invites the response "what else could you take on at that level?"

This framing also forces the discipline that protects you. If you cannot complete the sentence "I used AI to handle X, which freed me to do Y" with a real Y, you have a problem -- and it is better to find out before your manager does. The augmented version of your role is not a slogan; it is something you have to actually be doing.

What to avoid

Avoid making productivity claims that invite the question "so we could have one person doing two jobs now?" Do not show that AI can replicate your entire role without describing what you contribute on top of the AI output. Do not use AI output in client, board, or external stakeholder communications without reviewing it carefully -- if an error reaches a customer, it came from you, not the tool, and the trust cost is yours. Do not undersell your judgement: the temptation to attribute good outcomes to "the AI" rather than to your direction is real and is the wrong move politically.

A separate trap: avoid AI demos that make your role look like a button-press. Demos are not neutral. A manager watching you produce a polished output in three minutes will not internally model "I am watching three minutes of work" -- they will model "this role can be done in three minutes." Frame demos as showing the AI-assisted step inside a longer process, with explicit attention to the steps the AI cannot do.

Proposing the augmented version of your role

The most direct approach is to use your role audit (module 2) to propose a revised role definition to your manager. Identify the tasks you can automate or AI-augment, describe what you will redirect that time toward, and show the business logic. This is more compelling than waiting for your employer to redesign roles and hoping the result works in your favour. Most role redesigns done by management without employee input are blunter than the redesigns proposed by employees who understand the work.

A pre-conversation move that helps: float two or three of the redirected tasks informally before the proposal meeting. If your manager has already seen you do the higher-value work and liked it, the proposal is just formalising something already in motion. This is much easier to approve than a pure pitch.

If your employer is moving faster than you

If your organisation is already deploying AI on your tasks without a clear conversation about what your role becomes, that is a signal -- and the signal is usually that the organisation has decided where this is going but has not told you. Ask directly: "As we deploy AI on [task category], what does this role look like in 12 months? What can I do now to be valuable in that shape?" Getting a concrete answer -- even an uncertain or uncomfortable one -- is more useful than avoiding the question. Silence is not friendly here; silence usually means the answer is being figured out without you in the room.

If the answer is unsatisfying or evasive, that is information too. It does not necessarily mean the role is doomed; it does mean your employer is not actively planning to invest in your transition. That should weight your decision in the next module heavily toward switching or retraining rather than staying.

When to bring the union, the works council, or external advice

In Ireland and across the EU, AI deployment is increasingly covered by collective consultation rights. The EU AI Act and the Irish Workplace Relations Commission have specific frameworks for high-risk AI in the workplace, particularly around hiring, performance evaluation, and monitoring. If your role is in scope, your union or employee representative has both standing and resources you should know exist. This is not a confrontational move -- it is using a structure designed for exactly this transition.

Looking ahead

The next module covers practical tools you can try this month, with anti-hype framing. The final module covers the bigger decision -- retrain, switch, or stay -- once you have the audit, the skill investment plan, and the conversation framework in place.