Institutions are facing data cost increases of up to 50% next year as major vendors overhaul their pricing to account for AI, according to a new report from Substantive Research, BCG Expand and TRG Screen.
As vendors launch their own Model Context Protocol (MCP) servers and AI access models, many clients are being moved onto new commercial structures that can limit flexibility and create technical lock-in.
Speaking to The TRADE, Jane Fenton, global market data manager at Ninety One, said: “There’s some understanding that AI, MCP and APIs change the way providers need to think about licensing and protecting their IP, and in some cases the additional cost can be justified by better content or efficiency.
“The frustration is when clients are effectively forced into those models because legacy products are being replaced, leaving them with higher costs whether they want the new functionality or not.”
Read more – European traders keeping AI at arm’s length from execution
The report, from Substantive Research, BCG Expand and TRG Screen, found that neither vendors nor clients have yet settled on a standard approach to pricing AI.
Specifically, Substantive’s findings revealed that separate AI charges in vendors’ 2026 proposals ranged from $30,000 to $1 million.
This wide range suggests vendors are still determining how to price AI capabilities.
Index pricing has also drawn push back, with vendors offering average discounts of 34% in response to client resistance to new models.
However, despite these discounts, index product costs are still rising by nearly 20% year on year.
Speaking to The TRADE, Mike Carrodus, chief executive of Substantive Research, added: “New requirements are always an opportunity for vendors to reprice. But data consumers are struggling with the levels of repricing they are seeing in ratings, indexes, private markets data and alt data, in situations where their fundamental use cases haven’t changed, just the delivery method.
“This market was already struggling with the inconsistency of the application of pricing and pricing models from vendors, and the advent of AI has created another level of deep uncertainty on how to make sure you are optimising your spend.”
Industry market data spend is expected to rise a further 7% this year, according to BCG Expand’s H1 data, adding to the pressure from vendor price increases.
With agreements negotiated bilaterally and no shared pricing benchmark, firms have little way of knowing whether they have secured competitive terms.
As data costs continues to outpace clients’ budgets, the report warns that decisions made by vendors and clients today could set precedents for decades to come.