How LATAM is using Copilot to unlock more from service

AI adoption is no longer the finish line. As AI becomes a more established part of customer service, simply having AI “turned on” matters less. What matters now is how deeply teams put it to work.

The deepest AI adoption isn’t just about how much work AI can automate. It’s about how much more capable AI allows the operation to become. Our proprietary benchmark data offers a unique view into this way of thinking about AI maturity. And when we look across regions, one stands out: Latin America.

Of all customers with Zendesk Agent Copilot, LATAM teams have the highest usage of Agent Copilot per seat of any region. It also ranks first for the share of regional accounts using Zendesk AI Agents.

Taken together, those signals tell a bigger story.

Agents are building AI into their habits

One of the clearest signals of maturity is how often AI gets used in everyday work. LATAM teams use Zendesk Agent Copilot about 2x as much per seat as AMER, about 1.5x EMEA, and about 1.4x APAC.

There’s an important distinction here. EMEA has broader paid Zendesk Agent Copilot adoption, so this isn’t a story about which region has bought the most AI. It’s about what happens after adoption.

As agents use AI more consistently, it starts to change the way work gets done. Agent Copilot helps agents work through requests, find information, communicate with customers and move work forward more efficiently.

Instead of reaching for AI occasionally, agents begin to work with it as a matter of course. That’s when AI shifts from an added capability to part of the operating model. From there, AI’s impact can extend well beyond the service operation, changing what the business itself is capable of. The signals from LATAM suggest that’s the trajectory the region is already on.

More capable AI is giving service teams room to grow

Arajet brings that trajectory to life. The Dominican Republic-based airline uses Zendesk AI Agents and Agent Copilot to scale service as passenger demand grows. In aviation, that means handling time-sensitive requests while serving passengers across languages and markets.

AI Agents now resolve 70% of Arajet’s chat interactions without escalating to an agent, while chat represents 40–50% of the airline’s overall contact volume. 

At the same time, Zendesk Agent Copilot extends the team’s reach further. Real-time translation between Spanish, English, French, and Portuguese enables Spanish-speaking agents to support passengers across markets and languages. As Carlos Mesa, Vice President of Customer Experience at Arajet, puts it: “The [Zendesk Agent] Copilot tool has been fundamental for us.”

The result isn’t simply more automation, it’s support and added capacity for the human agents. In turn, this ripples out to enhance the entire service operation. 

“Zendesk AI customers are routinely achieving 80%+ automation rates.” Shashi Upadhyay, President, Products, Engineering and AI at Zendesk 

Arajet kept customer service costs and staffing levels stable despite passenger volume increasing roughly 40–50% year over year. Its agents now resolve 60% more tickets, while cost per interaction has fallen by 40%.

Arajet can now support more passengers across more markets without service costs growing alongside demand. Customer service becomes a growth enabler, creating the capacity for the business to scale.

LATAM offers a preview of what comes next

LATAM doesn’t lead every measure of AI adoption. But its performance across several measures of depth makes the region worth watching. More importantly, it gives us a clearer picture of what to look for as AI adoption matures elsewhere.

The next phase of AI adoption will be defined by depth: how naturally AI fits in and enhances to the way people work, how much responsibility organizations are willing to give it, and ultimately, how much more the business can do as a result.

LATAM is already giving us a preview of what that looks like.