GoldenSource and InvestOps Research Reveals Weak Data Foundations Are Putting AI Outcomes at Risk, Slowing Growth and Costing Investment Managers Billions
GoldenSource today announced the release of a new report with InvestOps Insights, The Basis Point Blind Spot: How Weak Data Failures Quietly Undermine Performance, Growth, and AI Ambitions.
Produced in partnership with InvestOps Insights and informed by 100 global buy-side firms, the report highlights a growing industry challenge: firms do not have an AI model problem, but a data infrastructure problem. The findings show that 98% of firms are concerned that weak data operating models are undermining AI outcomes, while 55% report 0.50 basis points or more of annualized exposure tied to AI and advanced analytics risk.
The findings suggest that low confidence in AI reflects a deeper issue in the data foundation. Fragmentation, weak governance, and lack of context make it harder to support trustworthy outcomes at scale. This is not simply a data quality issue, nor is it solved by centralizing records alone. Firms need to make relationships across capital markets data understandable. For example, they need to show how instruments, positions, counterparties, and legal entities connect so information can be interpreted in context and trusted across workflows and decisions.
“One of the most striking aspects of the research is who is feeling the pressure,” said Rory Pilbrow, Portfolio Director, InvestOps. “With 44% of respondents drawn from Portfolio Management and Executive Leadership, this is clearly no longer just a technology or operations issue. Financial loss, AI confidence, and the effects of poor data foundations are now being felt as firm-wide concerns.”
The impact of those same data weaknesses is also showing up more broadly across the business. The cost is not limited to direct financial loss. When data is inaccurate, delayed, or lacks context, the impact can compound across investment decisions, operational efficiency, regulatory oversight, and client trust. More than 80% of firms report financial losses linked to poor or late data, while 63% say their data model only partially or minimally supports a total portfolio view across public and private markets. The report also finds that 63% of Executive Leadership favor more significant change in the data operating model, underscoring the growing urgency around stronger foundations.
“It is encouraging to see Executive Leadership pushing for a faster pace of change in the data operating model. Firms that act now will be in a much stronger position to move faster, manage complexity, and apply AI with greater confidence,” said James Corrigan, Chief Executive Officer of GoldenSource.
GoldenSource will host a webinar on June 18 featuring buy-side leaders and GoldenSource executives to discuss the findings, how they are being interpreted in practice, and which changes firms should prioritize now.
GoldenSource is the Modern Financial Data Management Platform for capital markets, delivered through cloud-native infrastructure designed to ground AI in governed capital markets context.
For buy-side firms, it serves as an Investment Data Platform, helping clarify what they own, what it is worth and where they are exposed.
For sell-side institutions, it serves as a Trading, Risk and Regulatory Data Platform, supporting trading, counterparty exposure, valuation, risk, regulatory reporting, post-trade, settlements and control.
What sets GoldenSource apart is its Trusted Contextual Data Layer. It is built on the most extensive domain coverage, a structured financial data model, pre-configured data connections and deep buy-side and sell-side expertise. This turns fragmented data sources into a governed enterprise view with cross-domain relationship mapping.
The result is greater confidence in the information behind workflows, decisions and outcomes.
For more information, visit www.thegoldensource.com.
Media Contact
Chris Goodreds
Global Head of Marketing
GoldenSource LLC
[email protected]
+44 (0) 7350 421016
GoldenSource and InvestOps research finds weak data foundations are quietly undermining AI outcomes and costing buy-side firms across markets.