Three-quarters of enterprises have rolled back AI customer agents after deployment
Seventy-four percent of enterprises have shut down or rolled back an AI customer communications agent after putting it into production, according to research from Sinch AB released May 13, 2026. The rate climbs to 81% among organizations with mature governance frameworks.
The findings contradict the assumption that better oversight prevents failures. Instead, they suggest advanced monitoring lets companies spot problems that others miss.
"The most advanced organizations aren't failing less; they're seeing failures sooner," said Daniel Morris, Chief Product Officer at Sinch. "Higher rollback rates reflect better monitoring and control, not weaker performance."
The real problem isn't deployment
The research surveyed 2,527 senior decision-makers across 10 countries and six industries between January and February 2026. It found that 62% of enterprises already run AI agents in production-deployment is no longer the barrier.
The challenge has shifted to maintaining performance and reliability after launch. Engineering teams spend most of their time building and maintaining safety systems instead of improving customer experience, Morris said.
"Engineering teams are spending most of their time building and maintaining safety systems, a lot of which their communications infrastructure should be providing," Morris said. "That's the guardrail tax that slows organizations down."
Infrastructure matters more than governance alone
Eighty-four percent of AI engineering teams spend at least half their time on safety infrastructure. Yet governance investment alone isn't solving the problem.
Communications infrastructure satisfaction is the strongest predictor of successful AI deployment-stronger than investment levels or guardrail maturity. Eighty-seven percent of organizations rate high-performance infrastructure as essential or very important.
More than half of enterprises are building custom infrastructure to manage cross-channel context. Eighty-six percent have evaluated or are actively evaluating new communications providers.
Investment priorities don't match the problem
Enterprises invest more in trust, security and compliance (76%) than in AI development itself (63%). This makes governance the top investment category in AI programs.
Yet 98% of enterprises are increasing AI investment in 2026, suggesting they plan to expand customer communications agents despite current rollback rates.
The full report will be released in Q2 2026, with regional and industry breakdowns.
Your membership also unlocks:
