AI enterprise deployment remains in its early phases despite the technology’s potential to generate trillions of dollars in economic value, OpenAI chair and Sierra co-founder Bret Taylor said at MWC26 Barcelona on March 5, 2026.
Taylor made the remarks during a keynote session alongside Singtel Singapore CEO Ng Tian Chong. He noted that even if development of large language models stopped today, the economic benefit from existing artificial intelligence technology would still be substantial.
Major Industrial Applications Remain Largely Untapped
Taylor pointed to several sectors where AI adoption is still limited. Coding, software engineering, customer services, and legal services represent major application areas where most of the technology has yet to be deployed.
“If you look at the major industrial applications of AI; coding, software engineering, customer services, legal, most of the technology is yet to be deployed.”
Bret Taylor, Chair, OpenAI and Co-founder, Sierra
Taylor further stated that AI systems are evolving rapidly, with new models and increased adoption continuing to expand the technology’s capabilities. He added that observers three years from now would likely view the current moment as very early on the maturity curve.
AI Enterprise Deployment Must Start at the Top
Singtel Singapore CEO Ng Tian Chong described his company’s approach to AI adoption, emphasizing that leadership commitment is essential. Ng said he restructured the organization to ensure the AI initiative reported directly to him, rather than being buried several layers down the hierarchy.
“I did not want the AI initiative to be buried two or three layers underneath me, so I created a head of AI data and analytics reporting to me directly.”
Ng Tian Chong, CEO, Singtel Singapore
Data Architecture Is a Prerequisite for AI Agents
Ng also outlined the technical prerequisites for deploying AI agents effectively. He stressed that a solid data architecture must be in place before any agent deployment, because the agent’s performance depends entirely on the quality of the underlying data.
Moreover, Ng noted that businesses must treat AI not as a technology rollout but as a broader organizational transformation. He stated that companies positioning AI as an “organisation reinvention” are more likely to achieve lasting results.
Outlook: Organizational Change as a Driver of AI Success
Both executives agreed that the telecommunications and enterprise sectors are at an inflection point. However, the gap between available AI capabilities and actual deployment remains wide across most industries.
Consequently, the conversation at MWC26 Barcelona shifted focus from AI development to AI implementation — highlighting that infrastructure, leadership alignment, and organizational readiness are now the primary barriers to realizing the technology’s full economic potential.