AI for Good Geneva summit highlights Vivo's on-device AI security

The AI for Good Global Summit was held in Geneva from 7‑10 July 2026, organized by the International Telecommunication Union with more than 50 UN partner agencies and co‑hosted by the Swiss government. Over 12,000 participants from around 170 countries attended, making it a major venue for discussing the future governance of artificial intelligence.

Chinese smartphone maker Vivo presented a "Secure and Trustworthy Interaction Architecture for On‑Device AI Agents" serving hundreds of millions of users. The solution was selected from 267 applications submitted by 40 countries and included in the ITU’s “Innovate for Impact” case‑study collection. Vivo’s framework isolates the agent’s runtime environment, dynamically verifies intent, and enforces fine‑grained permission controls across infrastructure, system, application and human‑machine interaction layers. The company argues the approach reduces data‑privacy concerns, enhances user comfort and contributes to the UN Sustainable Development Goals.

Vivo representatives chaired workshop sessions on trust management for agentic AI, highlighting the company’s ongoing involvement in ITU‑T Study Group 17 on security standards. The summit also showcased a range of AI‑driven innovations, including robotics demonstrations, brain‑computer‑interface implants and locally‑processed AI medical systems, underscoring the broad relevance of trustworthy AI technologies.