The AI/ML salary band just snapped in two
The top of band hasn't moved — it has split. Two markets now operate side by side, and most benchmarks still price them as one.
Draft. Figures marked like this are illustrative and pending verification against Recruise placement data & Sachith sign-off before publication.
Key takeaways
- The AI/ML band hasn't inflated evenly. It has split into two — a broad market for applied ML and a thin, fast one for people who can ship GenAI into production.
- Benchmarks that report a single median for "AI/ML" now hide the number that matters. The spread inside the band is wider than the movement of the band itself.
- The hiring error isn't paying too little. It's pricing two different roles off one line, then losing the offer to a company that priced them separately.
One band became two, and most benchmarks still average them.
For a decade "AI/ML" behaved like a single, if fast-moving, band. You could take a median, add a premium for a hot skill, and be close enough to make an offer that landed. That is no longer true. The work has bifurcated — applied machine learning that improves an existing product on one side, applied GenAI that stands up a new capability on the other — and compensation has followed the work.
In our recent GenAI-adjacent mandates, the strongest offers cleared around ₹1.12 Cr, while the broad applied-ML market sat closer to its long-run median. The top of the band didn't rise so much as detach. Average the two and you get a number that describes no one you are actually trying to hire.
"The median stopped being useful the moment the band split. You're not benchmarking a role anymore — you're benchmarking two roles wearing the same title."
Sachith Rai · MD & Founder, Recruise
The cost of the mistake is an offer that dies in the last round.
When a GCC prices a production-GenAI leader off the blended band, the number looks defensible internally and light externally. The candidate has an offer from someone who priced the scarce version of the role, and yours becomes the one they use to negotiate — not the one they take. You spent the search and handed the close to a competitor.
The fix is upstream of the offer. Decide, at brief, which of the two markets the role actually sits in, and benchmark against that market only. It is a smaller, sharper comparison set, and it is the one the candidate is already using. When the split is this wide — on the order of a full band between the two — precision at brief is worth more than speed at offer.
One pattern worth knowing, every week.
The Signal is our weekly read on the senior GCC talent market — one chart, one pattern, no noise. Written from live placement data.
More from Recruise Insights.
The hidden cost of paying at the 75th percentile
It sounds defensible. It’s also why your offer-accept rate is slipping. The real benchmark isn’t where you think.
The published band and the closed offer are two different numbers
One is written by people not running mandates. The other is what the candidate actually signs. Only one of them should set your budget.
The AI/ML band didn’t shift — it split in two
Traditional ML sits where it was a year ago. The GenAI end has pulled away. Price them as one band and you lose the top while over-paying the bottom.