Singapore launches SG AI Xchange and asks companies to share what failed

Most government AI launches hand out money or rules. Singapore has launched something cheaper and harder. On 17 September 2026 its Ministry of Digital Development and Information published the opening address at the official launch of SG AI Xchange, a forum where companies tell each other how their AI projects actually went.[1]

That includes the parts that went badly. The address is explicit that the exchange is only as useful as what companies put into it, which means naming what proved difficult. For Malaysian bosses, the mechanism deserves more attention than the branding.[1]

The ministry dated the address 17 September 2026 and published it as the official launch of SG AI Xchange. It was delivered by Ms Jasmin Lau, Acting Minister for Manpower and Senior Minister of State for Digital Development and Information, who greeted the heads of the Singapore Institute of Technology, GlobalFoundries and the Economic Development Board.[1]

The address describes a sharing platform and community, not a fund, a regulator or a product. GlobalFoundries spoke at the first session about scaling AI in a manufacturing environment where quality, uptime and cost all matter. No fee, eligibility rule or meeting schedule is stated.[1]

The unusual part is the request. Ms Lau said the exchange only works if companies are willing to share what they have not done well, and granted that this is not common culture in Singapore. She asked leaders to name what they tried to solve and what proved difficult.[1]

She was equally blunt about the limits of a good conversation. Listening is only the first step. A session counts only if a pain point gets identified afterwards, a use case improves or work is reorganised. She told companies not to wait too long, because AI itself keeps changing.[1]

SG AI Xchange closed a busy week on the same ministry newsroom. On 15 September 2026 the ministry published Senior Minister of State Tan Kiat How's keynote at the inauguration of Google Cloud's Singapore Engineering Center, covering engineering and product development, multilingual and agentic AI, and cloud infrastructure.[2][4]

A day later it published Minister Josephine Teo's remarks at the Databricks Data and AI World Tour in Singapore. She named the impediments companies hit when they try to go deeper: a shortfall in skills, trouble integrating AI with existing workflows, and sometimes plain difficulty getting good data.[3][1]

A second theme runs through the address and it is not technical. Ms Lau described a group she said is rarely in the room: workers and younger students who are somewhat anti-AI, who fear for their jobs and who link AI to environmental damage. Those sentiments are real.[1]

Her advice to employers was direct. Do not take a quiet workforce for an agreeable workforce, because staff who are not pushing back are not necessarily content. If you plan to automate, tell people what could happen and what retraining suits them. If the outlook is weak, say so.[1]