Despite years of investment in AI and advanced analytics, most financial institutions are still fighting fires rather than preventing them.
That is the blunt assessment at the heart of a new fireside chat from SymphonyAI, which brings together two leading voices in financial crime and risk technology to examine why the compliance sector continues to lag behind — and what it will take to change course.
The conversation pairs Chartis Research principal analyst Sidhartha Dash with SymphonyAI Financial Services head of strategy and innovation Jason Shane. Dash is a globally recognised authority on risk technology, offering independent perspective on market trends, vendor capabilities, and the evolving regulatory environment. Shane leads strategic innovation across SymphonyAI’s AI-driven financial crime solutions, specialising in transforming compliance operating models through embedded AI and agentic workflows.
Together, they make a compelling case that the compliance sector’s struggles are not rooted in a shortage of capable technology. The tools exist. The problem, as SymphonyAI frames it, lies in how — or more often whether — institutions are actually putting them to work. Outdated operating models, siloed systems, manual processes, and sluggish change cycles are collectively preventing organisations from realising the value of the AI they have already invested in.
Central to the discussion is the concept of embedded AI — a meaningful departure from the conventional view of AI as a supplementary layer added on top of existing processes. In practice, SymphonyAI explains, embedding AI means integrating intelligence directly into the fabric of day-to-day compliance workflows. Alert enrichment, automated investigation steps, and real-time decisioning all become standard features of how compliance functions operate.
This embedded approach is what SymphonyAI and Chartis argue is needed to make the leap from reactive to proactive compliance. The traditional model — built on static rules and post-event investigation — generates excessive false positives, slows response times, and leaves institutions perpetually on the back foot. Agentic AI, by contrast, can detect emerging risks earlier, adapt continuously to new patterns, and support a more anticipatory approach to financial crime risk management.
The fireside chat also takes a candid look at where institutions most commonly go wrong when attempting to adopt AI. SymphonyAI points to a familiar set of pitfalls: viewing AI as a bolt-on fix rather than a structural transformation, neglecting the foundational work of data quality and workflow integration, misaligning governance frameworks with new AI capabilities, and pushing for rapid results without committing to a phased rollout. The discussion makes clear that incremental, modular adoption is not a compromise — it is the approach most likely to deliver lasting impact.
One recurring theme throughout is the balance between automation and human control.
For more insights, the full conversation is available via SymphonyAI.
Copyright © 2026 FinTech Global
