Abu Dhabi Global Market (ADGM) said it had completed the first phase of its artificial-intelligence rollout in a performance update published on 8 September. The financial centre reported deploying AI across 25 business functions, including licensing, supervision and customer service. It also said manual workload had fallen by more than 5,000 staff hours annually and that approximately 25% of customer enquiries were resolved instantly through digital channels. These are ADGM’s institutional performance claims; they are not presented as an independent impact assessment or as results for all public services in Abu Dhabi.

Other first-half figures in the same update put the asset-management ecosystem at 190 fund and asset managers and 276 funds, with 13,974 active licences. ADGM said institutions established in the centre held more than US$100 billion in AI-focused investment and that its own technology roadmap would involve more than AED400 million through 2029. Those numbers describe different things: capital managed by entities, accumulated measures of the business ecosystem, and a future technology investment plan. They are not interchangeable measures of spending or realised AI revenue.

The announcement said the reduction in manual workload exceeded 5,000 hours a year, but did not specify a baseline period, which tasks were measured, the calculation method or how staff hours were defined. The approximately 25% instant-response figure refers to customer enquiries through digital channels; it does not mean AI resolved every such enquiry. ADGM did not provide comparative values for approval times, error rates, customer satisfaction or total costs. The figures therefore indicate institutional use and reported operational outcomes, but do not establish the size of a causal effect.

For financial institutions, automation’s economic value depends on decision quality, security and oversight costs as well as response speed. ADGM said it intended to scale systems alongside data governance, cybersecurity, human oversight and clear authority boundaries, while retaining statutory decision rights with people. That matters because errors or opaque decisions in licensing and supervision can carry direct institutional consequences. Future reporting of processing time, repeat handling, corrected errors and complaints by transaction type would make the current claims easier to compare. Without that detail, the released measures provide a starting point rather than a complete assessment of productivity or service quality.