Infosys cuts FY27 growth guidance to 1-2% even as AI work reaches 10.5% of revenue
On October 23, Infosys revised FY27 revenue guidance to 1%-2% in constant currency, with margin guidance held at 20%-22%. That is down from 1.5% to 3% earlier in the year. The cut came in the same quarter it booked $5.2 billion in large deals, including a mega deal over $2 billion, and AI revenue grew 29% quarter on quarter to 10.5% of the total. CEO Salil Parekh said AI is giving growth while also causing compression. Its largest new deal contributes revenue only from the March quarter.
For a buyer, this is a pricing signal. Large vendors are winning work, and the same work is being delivered with fewer billable hours. Hiring has not stopped: India's five largest IT firms added a combined 13,918 staff in the quarter. But it is not tracking revenue, and your vendor knows its competitors are cutting prices. Our position is that time-and-materials rates for routine build, QA and maintenance work should be falling at renewal. This quarter, ask every vendor on a contract expiring in the next two quarters for a re-quote that assumes AI-assisted delivery, and move at least part of the scope to outcome-based or fixed-price terms so productivity gains show up in your invoice.
A question for your shortlist: Which AI coding tools does your team use on my account today, what share of delivery hours has that removed since last year, and how much of the saving shows up in my rate?
What Offshore Teams Get Wrong When They Review AI-Generated Code and Who Pays for It: AI writes code faster than offshore teams can review it. Here's what a real governance policy contains, five clauses to demand, and three controls to enforce.
React, Vue, or Angular: How Offshore Teams Are Actually Making the Call in 2026: Framework choice is now a staffing decision as much as a technical one. Here's how to match React, Vue, or Angular to the offshore bench you can actually hire.
The Offshore Vendor Tier Is Compressing and the Middle Is Getting Squeezed Out: Offshore vendors are splitting into AI-scale platforms and sharp specialists. The 50-to-500-engineer generalist is stuck between them, and buyers carry the risk.