Turo has activated a new agent within ChatGPT, enabling users to secure peer-to-peer vehicle rentals through natural language queries. This move signals a shift from simple chat interfaces to action-oriented AI capable of executing complex transactions.

Engineers should note the specificity of the function calling. Users specify constraints—pickup coordinates, dates, seat count, or powertrain type—and the model retrieves real-time inventory. In testing, a request for an electric vehicle with sufficient range for a 300-mile round trip returned viable options complete with charging infrastructure analysis. The system doesn't just list cars; it evaluates feasibility, noting Supercharger availability or specific charging speeds.

Unlike traditional fleets, Turo's decentralized inventory requires robust matching logic. The integration handles this by exposing vehicle metadata—photos, pricing, host ratings—directly within the chat interface before redirecting to the platform for final confirmation. Access requires enabling the Turo agent and using the "@Turo" trigger.

This implementation highlights the growing maturity of agentic workflows in commerce. Rather than navigating filters, users define intent. The backend must reconcile vague requests like "useful for moving" with structured vehicle attributes. As AI assistants become standard in automotive interfaces, from dashboards to return inspections, partnerships like this demonstrate how large language models can bridge the gap between unstructured human intent and structured database queries. The underlying architecture suggests a future where search engines evolve into execution engines, handling multi-step reasoning without manual intervention.

Source: CNET