Buy Side Research Spend Stagnates Despite 40% Surge in AI Tooling Adoption

Substantive Research has revealed that global investment research spending remains stagnant despite the rapid integration of artificial intelligence into buy-side workflows. For fintech professionals and asset managers, this disconnect highlights a critical tension: while AI-driven analytics are consuming a larger share of budgets, overall funding has failed to keep pace with the technological evolution of the sector.

What was announced

The latest survey from Substantive Research indicates that global research budgets increased by a mere 0.4% in the first half of 2026. This stagnation occurs even as the sell side struggles to monetize direct content feeds into client Large Language Models (LLMs). Geographically, the disparity remains stark; European research spending is currently 40% lower than that of North American counterparts. In the United States, a three-year recovery trend has reversed, with spending as a proportion of assets under management (AUM) decreasing by 6% this year, leaving American budgets more than 50% lower than 2018 levels.

The internal composition of these budgets is shifting rapidly toward technology. Spending on AI-enabled research analytics and tooling surged by 40% in a single year, now accounting for 7% of the average research budget. This growth comes at the expense of traditional providers. Concentration remains high, with the top 10 research providers capturing 56% of global budgets, while Independent Research Providers (IRPs) hold only an 8% share of the total spend.

Regulatory shifts in the UK and EU aim to address these funding gaps. Following the FCA’s COBS2 rules in July 2024 and PS25/4 in May 2025, asset managers can now charge for research alongside trading commissions for both segregated mandates and pooled funds. However, adoption of these Commission Sharing Agreement (CSA) models has been slow. The survey suggests that a significant switchover is not likely to begin in earnest until the second half of 2026, due to concerns regarding how asset owners will react to these costs being passed back to them.

"From a supply and demand perspective, the research industry is now rapidly transforming itself as AI changes the way fund managers want to consume research, and how providers create and deliver these insights. But until now, funding and budgeting for research has not evolved to accompany that evolution. The key question that needs to be answered is whether providers will accept that delivering this content directly into clients’ models is just another channel, or whether there is incremental value and cost that should be reflected in new pricing for this market."

Mike Carrodus, CEO of Substantive Research.

The companies involved

Substantive Research Ltd is a specialized provider of research and market data discovery and pricing analytics. The firm occupies a unique niche in the financial services ecosystem by providing transparency into how investment research is priced, consumed, and valued across the global buy side and sell side. Led by Founder and CEO Mike Carrodus, the company has become a primary source for benchmarking research spend, particularly in the wake of the MiFID II regulations that fundamentally altered how investment research is unbundled and paid for in European markets.

The firm’s data provides a granular view of the competitive landscape between traditional bulge-bracket banks, independent research houses, and the emerging class of AI-driven analytics providers. By tracking the flow of capital from asset managers to research producers, Substantive Research highlights the structural differences between the US market, which has historically maintained larger research pools, and the UK and EU markets, which are currently navigating a complex regulatory rollback designed to restore competitiveness and liquidity to local research environments.

What FF News has reported before

FF News has previously tracked the intersection of institutional research and emerging technology, specifically how regulatory and licensing hurdles impact the buy side. In July 2026, we reported that Broker Licensing Restrictions Stalling AI Adoption for 69% of Major Asset Managers. That report underscored the friction between traditional content delivery and the requirements of modern machine learning models. The findings highlighted that a significant majority of major asset managers were finding their AI ambitions throttled by legacy contractual frameworks, a theme that aligns with the current data showing the sell side’s ongoing struggle to effectively monetize LLM-integrated content feeds.

What this means

The flatlining of research budgets amid a 40% spike in AI tooling spend suggests a "cannibalization" phase within the fintech sector. Traditional research providers are under immense pressure; they are being forced to innovate their delivery methods via LLMs while their share of the wallet is being diverted to pay for the very tools used to process their data. The industry is currently in a holding pattern. Until the UK and EU fully embrace the return to CSA-funded models in late 2026, the funding gap between Europe and the US will likely persist, potentially leaving European asset managers with less sophisticated proprietary insights than their American rivals.