Chapter 5

Investing in trust isn’t enough

75% of organizations rank trust, security and compliance as their #1 AI investment priority – ahead of AI development itself – across every industry, region, and persona group in this study. And yet three in four enterprises globally have still had to roll back a live AI agent due to a governance failure.

Enterprises are deploying AI in customer communications to build customer trust, but the data shows it’s breaking it instead. The question is no longer whether organizations are taking governance seriously, but why taking it seriously isn't enough to correct the failure rate.

Read the chapter
Image for Investing in trust isn’t enough

Heavy investment, same failure rate

Enterprises invest more in trust, security, and compliance than in AI development. It’s the #1 spending category globally, selected by 75% of respondents, ahead of AI development at 63%. Organizations have voted with their budgets: Governance is the priority.

This consensus holds across industries. In tech, trust and security investment outpaces AI development by 10 points – 81% versus 71%. Healthcare (75%), financial services (78%) and retail (75%) also rank it first.

But our data shows this isn’t enough. 74% of organizations that reached production have still shut down or rolled back an agent. Among the most governed programs – the ones investing most heavily in safety – the rate is 81%.

Sinch data (2026) shows trust, security, and compliance is the #1 spending category in AI programs globally.

The AI Paradox phase ii - Anton

The guardrail tax compounds

84% of engineering teams are already spending at least half their time building guardrails and safety controls from scratch. Over a third spend most of their time there.

But the guardrail tax doesn’t flatten out as AI programs mature. It actually accelerates. Production-stage engineering teams – those that have already deployed AI agents – spend significantly more time on guardrails than pre-production teams. 38% of production-stage teams spend most of their time on governance, versus 29% of pre-production teams.

More agents, more channels, more compliance requirements – each one adds another layer that has to be built, maintained throughout the product lifecycle, and rebuilt when something breaks. The organizations furthest along are paying more for safety, indefinitely, for the same failure rate.

Where governance investment breaks down

Across the C-suite, trust is the one thing every leader agrees on. Every persona in the study puts trust, security, and compliance at the top of their AI investment priorities, ranging from 72% to 86%.

And yet when asked what’s actually blocking business impact, the same category comes back as the #1 barrier, cited by 37% of organizations – ahead of every other factor in the study. Budget constraints sit at just 7%. This isn’t a resourcing problem.

The category attracting the most investment is also the one blocking the most progress. When the most committed leaders and the most governed programs are failing at the same rate as everyone else, the problem isn’t who’s leading the program or how seriously they’re taking it. Everyone is trying to solve the right problem in the wrong place.

Sinch data (2026) shows trust, security, and governance concerns as the #1 barrier to AI business impact.

The AI Paradox phase ii - Daniel

The customer trust paradox

That mismatch has a cost, and it lands where enterprises can least afford it.

The primary goal behind AI communications investment is improving customer satisfaction and loyalty – cited by 36% of respondents, ahead of revenue and cost reduction.

But when AI fails in a customer interaction, the consequences land in exactly the same place. 34% cite reputational damage and loss of customer trust as the primary impact. What’s more: 46% of organizations identify customer data leakage as the #1 security risk in their AI programs. In production, at scale, that exposure doesn’t hit one customer. It hits thousands, and the trust it erodes doesn’t come back on the next deployment cycle.

When the investment designed to build trust is producing the failures that break it, more budget won’t close that gap.  

THE REGIONAL VIEW

The governance gap is a global story

Trust, security and compliance is the #1 investment priority globally. From India at 86% to France at 64%, governance is cited as a main investment area across all countries in the study. The commitment is universal and consistent.

So are the failures. Across every region in the study, at least six in ten AI communications programs have experienced a rollback. In some markets it runs as high as nine in ten. No region has cracked the governance problem, regardless of how much they’re investing in it.

The global consensus on governance investment is clear. What’s less clear is whether it’s going where it needs to.

Sinch data (2026) shows trust, security, and compliance as the number one investment focus across countries.

Image for What to do with this

What to do with this

Trust, security and compliance attracts more investment than any other category in AI communications. And yet 74% of organizations have still rolled back a live agent due to a governance failure. The most governed programs fail even more.

Start with your own numbers. Compare your governance investment trajectory against your rollback trajectory. If the spend has gone up and the failure rate hasn’t come down, the cost is compounding. And the next budget increase will do the same.

The next chapter looks at why, and what the organizations closing the gap are doing differently. 

The visibility gap that’s dividing leaders Chapter 4 The visibility gap that’s dividing leaders Infrastructure is where the race gets decided Chapter 6 Infrastructure is where the race gets decided