Tern launches agentic AI to automate travel advisor workflows

Tern’s agentic AI rollout is a good example of how ‘boring’ workflow automation is quietly absorbing a lot of the gains from frontier models. Instead of pitching a general chatbot, Tern is wiring model‑driven agents directly into the daily grind of travel advisors: reading inboxes, assembling proposals from live inventory, and drafting replies that feel like the human advisor. The human‑approval step is crucial—it keeps liability manageable while still letting the AI do most of the cognitive heavy lifting.

The strategic shift here is from AI as a tool users call explicitly to AI as an ambient co‑worker that touches every transaction. In sectors like travel, where margins are thin and coordination costs are high, that can change the economics of small agencies and host networks. If one platform can reliably offload 30–50% of white‑collar work to agents without breaking trust, that pattern will be copied in law, insurance, consulting and beyond. Those demand‑side pulls support continued investment in more capable, longer‑horizon models.

In AGI terms, Tern is not moving the research frontier, but it is increasing the surface area where near‑AGI systems can be tested in production. The more businesses normalise agentic workflows, the more data and revenue flow back to the labs building sophisticated planning and tool‑use capabilities, which indirectly nudges the frontier forward.