Genpact: 79% run AI agents, 15% tie them to P&L

Big companies are running AI agents in production. Most can't show you the profit. A new Genpact report, Autonomy in Practice, says 79% of the enterprises it surveyed have AI agents live, but only 15% have fully tied AI investment to P&L outcomes. If companies with whole AI teams can't measure the return, a 10-person shop needs to measure from day one.

What actually happened

Genpact published the report on November 12. The report page says it surveyed 2,005 enterprise executives. The release says they come from 17 industries, at companies with $1 billion to more than $50 billion in annual revenue. The answers are self-reported. The numbers:

• 79% have AI agents in production. Only 20% use them broadly and consistently across functions.
• 15% have fully tied AI investment to P&L outcomes. (The report page words it as 15% of AI initiatives.)
• 49%: on average, the share of AI initiatives funded in the past 24 months that delivered the expected value.
• 9% have finished a significant operating-model transformation.
• 40% of decisions are now led or fully run by AI, more than double a year ago. Humans still decide or audit in at least 87% of cases at every decision stage.
• 92% of respondents with agents in production kept their headcount the same or grew it.

Why it matters for your business

This is a survey of billion-dollar companies, so don't copy their numbers onto your shop. Copy the lesson. Getting an agent running is the easy part. The hard part is proving it paid for itself.

Big companies can absorb an AI project that falls short of plan. A small business usually can't. Before you build or buy an agent, write down three things:

1. The baseline. Hours per week the task takes now, and who does it. Measure it, don't estimate it.
2. The P&L line it should move. Labor cost, refunds, ad spend, missed calls turned into bookings. Pick one.
3. The kill date. If the number hasn't moved in 60 or 90 days, turn it off.

Then keep a person in the loop where it matters. Genpact's own respondents still have humans deciding or auditing in most cases. That is not a failure of AI. That is how you keep control.

The 92% number matters too. In this survey, agents mostly added capacity rather than replacing staff. Plan your agent around what your people do with the hours it frees up.

Sources: Genpact via PR Newswire, Genpact, Autonomy in Practice.