Ant Group CEO Han Xinyi: Enterprise AI can’t stop at boosting individual efficiency—build organizational intelligence

robot
Abstract generation in progress
Mars Finance News On July 20, during WAIC 2026, Ant Group CEO Han Xinyi shared three thoughts on how enterprises can deeply implement AI: intelligence is not limited to accumulating knowledge, but lies in the sedimentation of experience; value is not only improving individual efficiency, but also in organizational evolution; and the path to implementation is to build “Institutional AI (organizational intelligence)”. He cited the “Age of Experience” assertion by Sutton, a Turing Award laureate, arguing that general-purpose large models are hard to understand the enterprise’s operating logic. Enterprise-grade AI must take root in real business processes, accumulate proprietary data, and form Experience AI (experience-based intelligence). Han Xinyi also said plainly that most enterprises’ intelligent agents today still remain at the tool level—things like looking up information and writing meeting minutes—while existing processes are “designed for people” rather than “designed for agents,” resulting in everyone doing their part without anyone being accountable for the outcomes. He provided an implementation roadmap: first, embed agents into key workflows to accumulate experience—without proprietary data, there is no proprietary intelligence; then build a trustworthy data foundation using blockchain and confidential computing; finally, create a closed-loop of experience to drive an overall upgrade of organizational productivity. Meanwhile, Ant Digital Technology released a “super factory for intelligent agents” plan. (Wide-angle observation)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned