Wall Street’s Data Hunger Is Growing and AI Is Making It Riskier 

Hedge funds and private equity firms have long sought an edge. Now many of them are using artificial intelligence to mine vast troves of unusual data — everything from satellite images of store parking lots to streams of credit card transactions — in search of an investing advantage before anyone else spots it. The strategy is working. But it’s also getting more complicated.

That’s the central finding of a new report from Lowenstein Sandler, a national law firm that has tracked how investment managers use so-called “alternative data” since 2019. The firm surveyed private fund managers (including hedge funds, private equity firms, and venture capital firms) and the results show a dramatic shift in how this industry operates.

The numbers are striking. Nine out of ten respondents said they currently use alternative data. That’s up from 67 percent just last year, and 62 percent in 2023. Two years ago, it was a tool used by a minority of firms. Now it’s nearly universal. The alternative data market itself has grown to an estimated $15 billion globally.

What’s driving the surge? AI. Investment managers are increasingly using AI systems to process and analyze these massive, fast-moving datasets — the kind that would be impossible for human analysts to sift through manually. The technology is helping firms find new patterns and signals that point toward profitable investments, a capability the industry calls seeking “alpha.”

Read more: Pentagon’s AI Push Faces Friction With Anthropic Over Usage Restrictions

The report notes that AI is proving effective at “surfacing new signals, streamlining research, and unlocking new sources of alpha from complex, high-velocity datasets.”

But the same power that makes AI useful also makes it dangerous. The report warns that the technology is simultaneously “heightening firms’ exposure to model risk, governance gaps, and evolving data rights.” In plain terms, when an AI system makes a bad call, it can be hard to figure out why. And when firms feed AI with data that was obtained improperly or used without clear legal rights, they face serious legal exposure.

Those concerns are reflected in how much firms are spending. More than two-thirds of survey respondents said their alternative data budgets now exceed $1 million. That kind of spending signals commitment — but it also means the stakes are high if something goes wrong.

The report concludes that firms need to implement “disciplined controls, auditable decisioning, and a clear data provenance strategy” to turn this innovation into lasting returns. That last term — data provenance — refers to being able to trace exactly where your data came from, how it was collected, and whether you have the legal right to use it. It sounds like a technical detail; increasingly, it is a legal necessity.

Only 4 percent of respondents said they do not expect to use alternative data in the future. That means the regulatory and compliance questions raised in the report are not going away. The full Lowenstein Sandler analysis digs deeper into how firms are managing model risk, what governance frameworks are emerging across the industry, and how regulators are beginning to pay closer attention to these practices.