FINDINGS FROM 2,500+ ENTERPRISE LEADERS
Find out why getting to production wasn’t the hardest part.
See what the rollback data actually says about AI at scale.
Understand where the cost of failure really lands.
Discover how leaders report different realities of the same AI program.
See how more governance spend isn’t buying fewer failures.
Find out why investment alone isn’t closing the gap.
See who participated in the research and get the survey methodology.
For two years, the dominant story around enterprise AI has been about being stuck in the pilot phase. In customer communications, that story is no longer true. Sinch research (2026) shows 62% of enterprises already have AI agents in production. The debate about whether to deploy is over, but what happens next isn’t what the market expected.
Nearly three in four enterprises have been forced to shut down or roll back a deployed AI communications agent. And that number doesn’t decline with experience. Among organizations with fully mature guardrails, it’s higher. This is what production actually looks like.
When an AI communications agent fails, the impact splits three ways: the support queue, the brand perception, and the engineering team. Most organizations are tracking the first, but the other two are where the real cost accumulates.
Leaders across every role and department are looking at the same AI challenges and seeing different things – from rollback rates to the cause of failure and the engineering burden that goes with it. That gap shapes their confidence in their AI readiness and defines priorities. And that’s what’s putting AI programs at risk.
Enterprises are investing more in trust, security, and compliance than in any other part of their AI programs, and yet it’s still the #1 barrier to AI business impact. Organizations globally are prioritizing governance but rollback rates aren’t declining. And what’s worse: failures are directly impacting customer trust, the one outcome every leader is trying to protect.
Infrastructure satisfaction is the strongest predictor of AI deployment success across every variable analyzed, and yet for most enterprises, it’s the clearest gap in their current setup. Investment in AI communications is growing, but can the infrastructure underneath handle what’s coming?
2,527 enterprise decision-makers. 10 countries. 6 industries. We surveyed the leaders responsible for their organization’s AI communications strategy. Here’s how the research was conducted and what the methodology covers.
Here’s what 2,527 enterprise decision-makers reported, and what it means for organizations deploying AI in customer communications right now.
“The industry has assumed that better governance leads to better outcomes. But that’s not enough: If governance was the fix, the most mature teams would roll back less, not more. Our data points to a deeper issue. Engineering teams are spending most of their time building and maintaining safety systems, a lot of which their communications infrastructure should be providing, instead of focusing on improving the customer experience. That’s the guardrail tax that slows organizations down.”
Sinch Newsletter
In Production is a monthly newsletter for technology leaders who want to know what’s actually happening behind the market’s confidence numbers, written by people who know what production really looks like.
