Data readiness and quality are the #1 success factor for implementing GenAI (42% of respondents ranked this as their top factor), according to a study by Celent, a GlobalData company.
Janey Speed, Capital Markets analyst at Celent, says: “For AI use cases to not only be effective but also robust, firms must ensure that they are training their models with high-quality data. With the internet full of questionable AI-generated content, newer general models run the risk of being trained on bad data and poor responses.”
Implementation and production of agentic AI
The report titled, GenAI-oneers in Capital Markets, 2025 Edition, also revealed that financial institutions across capital markets indicate respondents are considering their expectations around GenAI and Agentic AI.
Financial institutions have swiftly implemented use cases, such as summarisation, knowledge sourcing, and process automation. However, the current focus is on urgently developing transformational, revenue-generating use cases.
As for the implementation and production status of agentic AI, the study found that only 26% of respondents have put it into production. However, by the end of 2026, more than half (52%) of participants anticipate rolling out an Agentic AI use case within their organisation.
“This data shows that capital markets firms must ensure that they are developing their tech strategy with the assumption that Agentic AI workflows are extremely achievable and likely in the not-so-distant future,” Speed said.
Celent surveyed respondents from 23 firms across investment management and securities trading, including buy-side, sell-side, and financial market infrastructure firms.
.
