The Number That Changed
A new report from Cambridge Judge Business School found that seven in ten investment managers are actively deploying AI in their front offices — up from roughly one in ten the year before. That's not a gradual adoption curve; that's a step function. The 2026 Global AI in Financial Services Report, drawing on responses from firms across asset management, private equity, and trading, found that the shift accelerated as firms moved from experimenting with general-purpose tools to deploying purpose-built workflows for portfolio analysis, earnings research, and risk monitoring.
What Front-Office AI Actually Looks Like
The most common use cases are earnings transcript summarization, portfolio risk briefings generated before market open, and compliance-ready documentation for investment decisions. Less common but growing: real-time signal generation from alternative data feeds and AI-drafted memos that human analysts review rather than originate. Firms with agentic AI in client-facing workflows saw 5–8% revenue increases and 15–20% jumps in customer satisfaction — numbers that, once visible to competitors, compress the time horizon for holdouts who are still in pilot mode.
Fintechs Are Further Along
Fintechs lead traditional institutions on agentic AI adoption: 57% versus 45% deployment rates. This gap matters because fintechs move faster from pilot to production — they don't carry the compliance approval cycles that slow institutional deployments — and they're setting expectations for what AI-assisted investment products should feel like. Incumbents are being judged against an experience baseline built by firms without legacy infrastructure constraints. That's a structural disadvantage that compounds over time.
What Builders Should Take From This
The 70% adoption figure masks a wide quality spectrum. Most front-office AI today is summarization and document review — valuable, but increasingly table stakes. The firms capturing outsized returns are running multi-step agent workflows: ingest earnings call, cross-reference filings, update internal model, surface flag to PM. Building infrastructure or products that plug into these workflows rather than replace them wholesale is the architecture that clears procurement. The report also noted that innovation overtook operational efficiency as the primary driver of technology investment for the first time in three years — budget is shifting toward differentiation, not cost reduction.
The Road Ahead
81% of respondents expect agentic AI to be meaningfully deployed across their organizations by 2030, with near-term focus on the shift from AI as assistant to AI as workflow participant. The firms that define this category won't be those with the best model access — that's a commodity. They'll be the ones with the cleanest data pipelines, tightest domain-specific toolchains, and the operational discipline to run AI in environments where a confident wrong answer has real consequences.