Last week a manager walked into my cabin and said, almost casually, “Sir, AI ne mera 40-50% kaam kar diya. Appraisal mein toh consideration milega na?”

I didn’t reply immediately. I just looked at him.

I suspect this conversation is happening quietly in a lot of companies right now. Not in townhalls. In corridors, in one-on-ones, and inside appraisal forms where people suddenly start talking about “efficiency” like it’s a new skill they developed.

Most of us in HR have already seen the shift. The person who used to take three days on a presentation now finishes it in under an hour. The recruiter who screened resumes one by one is now just reviewing shortlists. The operations guy who spent half his day making trackers is mainly checking what the system has already prepared.

Then appraisal season comes.

And the same people expect the same 10-12% they used to get when they were doing everything themselves.

The organisation is looking at it differently. If the same headcount is delivering more work with less effort, why should the increment percentage stay untouched? I’ve heard this question in closed rooms. One senior leader put it quite bluntly: if the job has become easier, why are we still paying the same premium for it?

That’s an uncomfortable sentence. Once you accept the logic, the entire appraisal conversation starts changing. It’s no longer just about effort or loyalty or “not creating problems.” It becomes about what value is still left after the AI has finished its part.

Most employees aren’t ready for that framing. They still operate on the old Indian workplace contract — you show up, you complete the work, you don’t resign, you get your 8-10%. That contract is under pressure now.

I saw a version of this play out in a mid-sized retail setup. A commercial team of six had started using AI heavily for planning, competitor checks, and even first drafts of vendor communication. Their output looked strong. During appraisals, a few of them expected above-average hikes. The business head simply asked what would happen if the same work was given to four people next year with better tools.

No one had an answer.

This is the part we in HR are still dancing around. AI isn’t only changing how work gets done. It’s changing the power balance across the appraisal table. The employee feels they delivered more, so they should get more. The organisation sees that more was delivered with less human effort, so the role itself may be worth less than before.

Both sides have a point. Both sides are also slightly hiding something.

Employees rarely admit how much of the heavy lifting the tools are doing. Organisations rarely admit they still need the human parts that remain messy — judgment under pressure, reading the room, handling politics, and dealing with the undefined 20% of work that no tool currently handles well.

The 10% number isn’t really the core issue. The bigger problem is that we’re still using old appraisal language for a new reality. We keep rating people on ownership, proactiveness and delivery without properly redefining what those words mean when a large portion of the delivery is machine-assisted.

Some companies will start differentiating quietly. People who use AI well and still bring clear judgment, faster decision-making and real stakeholder management will continue getting healthy increments. People who mainly become efficient operators of tools will see the percentage come under pressure.

It won’t be announced anywhere. It will just start reflecting in the final numbers.

And when employees notice, the next question will come: “Sir, AI use karne pe punishment mil raha hai kya?”

The people who will still comfortably get that 10% (and more) are the ones who can sit across the table and say, without performing, that AI did the 50%, but they did the part that actually moved the needle — and then show it.

AI may make your work faster. It won’t automatically make your contribution more valuable.

Leave a Reply

Discover more from Mayankkulshrestha.com

Subscribe now to keep reading and get access to the full archive.

Continue reading