Citi Securities Services Evolution 2025

32 | Securities Services Evolution 2025 Gen AI: From client onboarding into the back office “Everyone” is using it – but not in post-trade Given the significant volumes of industry discussion that has been dedicated to Generative Artificial Intelligence (GenAI) in the last year, it is no surprise to learn that 86% of respondents in our survey this year claim to be running GenAI pilots in their firms, with 53% apparently using GenAI in a live environment. Even allowing for potential hype-driven bias in these numbers (as we have seen with DLT adoption figures for years), there is no doubt that the intent to use GenAI across the financial institution is today real and meaningful. The only thing is that this adoption is taking longer to reach the back office, with only 57% and 24% of respondents piloting and live on GenAI in the post-trade space today. In 2025, GenAI’s primary use appears to be in bringing new answers to an age-old problem – client onboarding . Across the trade cycle, 83% of brokers, 63% of custodians and 60% of asset managers are using GenAI to make a meaningful impact in an area that has so far defied all attempts at automation (Figure 17). Where previous initiatives have failed to gather momentum, the ability of GenAI to increase Figure 17: Gen AI is an opportunity for client onboarding for all segments Question: In which areas are you running AI projects today – and at what level? Expressed as: % of respondents, per segment and per activity type. Institutional investors Asset managers Custodians Banks Broker/ dealers IT development 58% 25% 32% 50% 50% Client onboarding (KYC) 50% 60% 63% 29% 83% Issuance workflows 50% 30% 42% 36% 67% Product distribution/sales 50% 40% 42% 21% 33% Portfolio modeling 58% 45% 32% 50% 50% Surveillance and risk monitoring 58% 50% 37% 36% 50% Trade lifecycle management 67% 40% 42% 43% 50% Asset servicing 58% 45% 37% 36% 33% Post trade reconciliation & reporting 67% 40% 47% 43% 67% 32% 39% 25% 56% 23% 43% 20% 41% 20% 48% 27% 47% 13% 47% 16% 41% 16% 48% 53% Overall Production 86% Pilot % of respondents with deployments ongoing for Generative AI