Copy article

Morningstar Plugs Investment Research Into Perplexity: Trusted Data, Unregulated Delivery?

ended 11. May 2026

Morningstar and PitchBook have expanded their integration with Perplexity, an AI search tool that answers questions in conversational plain English instead of returning a list of links. That now includes questions about funds, markets, and where to put your money.

The deal pipes analyst-backed investment research directly into Perplexity's chat interface through a live data connection, so the AI draws on Morningstar and PitchBook intelligence when users ask investment questions. 

The press release frames this as "trusted intelligence in AI-powered research workflows." What it actually describes is a financial data titan embedding itself as the default source inside conversational AI, the place where a growing number of investors now form their first impressions before opening a Bloomberg terminal or calling an adviser.

AI in financial advice is a hot topic. Financial data lends itself to machine analysis, but when the output arrives as a conversational answer rather than a data table, mistakes get dressed up in confident prose and become very hard to spot. AI models aim to please the user and have a well-documented tendency to fill gaps with plausible-sounding fabrications rather than admit uncertainty.

The FCA's Mills Review is examining how AI will reshape retail financial services by 2030. The Treasury Select Committee warned in January that regulators' wait-and-see posture on AI in finance risks serious consumer harm. Meanwhile, a study published this week by US communications firm 5W and wealth-planning publisher Haute Wealth found hallucination rates on financial citations running between 20% and 37% across major AI models. 

Against that backdrop, Morningstar is positioning its proprietary data as the antidote to unreliable AI, but the integration itself still runs through a system with known accuracy problems, and sits outside the FCA's regulatory perimeter for financial advice. 

“Our focus is on delivering independent, analyst‑backed intelligence in ways that align with how investors and financial professionals work today,” said Adam Wheat, head of Data & Research Solutions, chief technology officer for Direct Platform at Morningstar. “By making Morningstar and PitchBook content available in Perplexity, we’re extending the reach of our data and research while maintaining the rigor investors require to act with confidence when it matters most.”

We'd like your views:

  • If AI-surfaced investment research leads to a consumer loss, does liability sit with the data provider, the AI platform, or the user who trusted a chat interface with their pension?
  • Morningstar manages approximately $370 billion in assets under management and advisement. When the same company supplying the data also manages the money, does embedding that data into AI search create an undisclosed conflict of interest?
  • The FCA has said it won't introduce AI-specific rules. If proprietary financial data becomes the default "grounding source" inside AI tools that millions use, is principles-based regulation adequate, or does this need a perimeter conversation now.

4 responses from the Newspage community

Copy all

Star Quote
Copy

Colette MasonVerified
AI Ethics Consultant at Clever Clogs AI
Better data going into AI is obviously a good thing. But when one company supplies the research, manages $370 billion in assets, and now sits inside the tool people use to ask "what should I invest in," that's not just a data integration. It's a new kind of gatekeeper. The investor doesn't see a Morningstar logo and think "that's a commercial data provider with its own funds." They see a confident, well-sourced answer in plain English and assume the search engine found it impartially. That's a very different relationship to information than opening a Bloomberg terminal or sitting across from an IFA, and right now nobody's regulating the gate.
Copy

Editor
Katrina YoungVerified
Futurist | Chief Technology Officer CTO/CDO Ethical AI & Digital Transformation Strategist at KYC Digital
Morningstar embedding proprietary data into Perplexity is being framed as a solution to AI reliability in finance. It is not. It improves one layer, data sourcing, while leaving the delivery mechanism largely unaccountable. The reported 20% to 37% hallucination rates on financial citations are not simply a data issue. They are a system behaviour issue. High quality inputs do not stop AI models from presenting uncertainty as confidence or generating plausible sounding answers from partial information. The structural gap is governance across the full chain: data provenance, model behaviour and conversational delivery. Consumers experience this as one trusted answer, while accountability sits across separate entities. Regulators may eventually view that fragmentation as a governance failure, not an innovation gap.
Copy

Editor
Mitali DeypurkaysthaVerified
AI Governance Strategist & CMO at AISensa
More than 28 million Brits are already using AI to help manage their money, according to Lloyds Banking Group, but we are now drifting towards a world of chatbot cash tips without clear accountability. Britain has a long and painful history of financial mis-selling scandals, yet at least consumers usually knew who to fight when things went wrong: a bank, a broker or a named adviser. If AI search becomes the new shadow financial adviser, the danger is that the blame simply gets bounced around: data firms say they only supplied information, AI companies say they only summarised it, and the consumer is left bearing the loss. I still remember my own father battling for compensation after being mis-sold an endowment mortgage in the 1980s, and I genuinely wonder whether someone in that position today would even get their day in court if the bad advice came from a chatbot instead of a human being.
Copy

Rohit Parmar-MistryVerified
Founder at Pattrn Data
This is not just better search. It is financial influence with a friendlier interface. Once investment research is delivered as a confident chat answer, consumers stop seeing a source, a model, and assumptions. They see a recommendation-shaped paragraph.

The liability question cannot be waved away by saying the user should have known better. If a data provider supplies the intelligence, an AI platform packages it, and both benefit from trust in the output, both need accountability when the answer is wrong, incomplete, or conflicted. The grey area is the product.

The conflict point matters too. When the same ecosystem can inform the answer and manage the money, disclosure cannot live in small print. In AI Audits, we look for traceability, uncertainty, and human escalation before automation touches high-stakes outcomes. Finance needs the same discipline. Principles-based regulation is useful, but only if the perimeter is honest. A pension is not a search query. It is someone's future.