Past the pilot: who is actually buying AI in finance. Qaike Team Published on: 27 Oct 2026 Surveys say almost every fund manager now uses generative AI. AIMA puts the figure at 95%, up from 86% in 2023. Yet between late April and late October 2026, Qaike's vendor news archive logged only about a dozen deals where a named firm said it had chosen an AI tool. This post looks at those deals, at the banks and asset managers that bought stakes in the vendors they use, and at the case studies vendors publish to fill the gap. Read together, they show where AI has moved past the pilot stage, and where it has not. Three kinds of signal. Not every "client win" carries the same weight. | Signal | What it is | What it proves | | Named selection | A firm says, on the record, that it has chosen a vendor | A real commitment, often beyond one team | | Strategic stake | A bank or asset manager invests in a vendor it uses | The buyer wants a say in the product, not just a licence | | Case study | The vendor describes one client's results | A use case works for one team, as told by the vendor | Signal 1: the named selections. The largest private equity name in the period is Clearlake Capital, which manages more than $185 billion. In July it partnered with Databricks and West Monroe to build an AI data platform across the investment lifecycle. In August it signed a deal with OpenAI to bring its models to more than 50 portfolio companies. Databricks is a general data platform rather than a finance specialist, which says something about where large firms start. Other investment firms chose specialist tools: * HB Wealth, a $30 billion wealth manager, chose Arch to run its private markets operations and reporting. * Brandon Capital, a venture firm, chose Quantium for fund accounting, portfolio management and investor reporting. * BSQ Capital Partners joined 73 Strings for valuation work in Asia-Pacific. The clearest firm-wide private equity deal this year predates its window: H.I.G. Capital, with $74 billion under management, picked ToltIQ in March after a two-month trial. Banks and advisers announced more often. AlphaSense says six more of the world's ten largest investment banks chose its platform in Q3, along with more than 10,000 new banking users. It now serves 18 of the top 20 banks, but names none of them. Standard Chartered partnered with Unique on investment intelligence for its wealth clients. Intapp named four buyers: investment banks Republic Partners and Ensis Partners, and M&A advisers Forvis Mazars Capital Advisors and Carr, Riggs & Ingram. All four use DealCloud to run deal pipelines and client relationships. Intapp says about 1,700 private capital and investment banking firms use its software. Several more deals left the buyer unnamed. A $12 billion North American PE firm took on Deal Engine for sourcing. A Fortune 500 asset manager and a Fortune 500 insurer chose Alkymi for private credit operations. Over six months, the pattern is clear. Nearly two in three buyers Qaike Ltd could count were investment firms: private equity, venture, credit, wealth and asset managers. But most of them appear only in case studies or without a name. Banks and advisers are fewer, and more likely to go on the record. No hedge fund appears at all, named or unnamed, which fits the sector's habit of secrecy. Signal 2: Buyers on the cap table. The stronger signal may be money rather than contracts. Four vendors took strategic stakes from the kind of firms that buy their tools. | Vendor | New strategic backers | Date | | Rogo | Barclays, BNP Paribas, Citi Ventures, MUFG Innovation Partners, Société Générale | September | | Finster | UBS Investment Bank, alongside FactSet | August | | Arch | MUFG Innovation Partners, Franklin Templeton | July | | F2 | Golub Capital ($5 million, plus a platform deal) | September | The Rogo round joins existing backers J.P. Morgan and Truist, taking it to nine banks on its cap table. Dealroom puts the round at about $30 million. Rogo's reach has grown fast: from more than 35,000 users at over 250 firms in April to more than 50,000 at over 350 by September. Finster's round is part of its Series B. Arch says it now has more than $500 billion of assets on its platform. F2's deal with Golub pairs a stake with a contract. MUFG backs both Rogo and Arch, one built for bankers and one for allocators. Service firms are buying in too: Broadridge took a minority stake in CENTRL in April, and Accenture invested in AlphaSense in June. A stake gives the buyer a voice in the roadmap and an early look at new features. For the vendor, it is a reference that no case study can match. For everyone else, it raises fair questions. If a rival bank owns part of your research tool, whose needs set the product plan? And who can see what? Signal 3: Case studies, and why the small print matters. Many firms use an AI tool without ever saying so in public. Vendors fill that silence with case studies: their own account of how one client uses the product. In this period, case studies covered eleven use cases, most of them in deal sourcing and diligence: * Finding targets and buyers: Grata with Sancovia, Ridgefield Partners, Mangold and TAE Europa. * Screening companies and deals: Grata with Quarza Inversiones, and V7 with 10 East and Moderne Ventures. * Underwriting new deals: Mosaic with Riverside. * Producing macroeconomic research: Macrobond with T. Rowe Price. * Reporting to investors: Quantium with Project A. * Closing the books and tracking portfolio data: Brownloop with an unnamed $20 billion mid-market PE firm and an unnamed global buyout firm. * Preparing for investor meetings: CENTRL with an unnamed $40 billion multi-asset manager. * Running daily investment operations: FINBOURNE with OMBA. * Researching fund managers: DiligenceVault with GEM Investments. * Managing research notes: Bipsync with Cherry Creek Family Offices. * Building firm-wide AI systems: Brownloop with an unnamed mid-market growth equity firm. Case studies are useful proof that a use case works. They are weak proof of scale. The vendor writes them, picks the client and sets the baseline, and claims such as "three weeks to 30 minutes" are rarely checked by anyone else. What do industry surveys say? The public deals match what surveys say about how firms use AI. Mercer's 2026 survey of 131 asset managers found 55% have AI in at least one investment process. Most use it to improve operational efficiency, such as automating routine tasks (73%), or as a partner that provides insight and analysis (68%). Only 5% give AI any decision authority, such as trade or investment recommendations. Use falls to 13% in portfolio construction and 7% in trade execution, and just 8% report better returns from AI. Private capital points the same way: AI is used far more for support work than for decisions, though Affinity's 2026 survey found the share of PE and VC firms using AI for investment decisions doubled to 28%. L.E.K. found real productivity gains but little deployment at scale. About the surveys * AIMA with Marex (2025): about 150 hedge fund managers, asked whether and how they use generative AI. Counts any use, however small. * Mercer (2026): 131 asset managers worldwide, surveyed online in February 2026, on how AI is used across the investment process. * Affinity (2026): about 275 private equity and venture capital investors, on their outlook for the year and their use of AI. * L.E.K. Consulting (2026): 100 US buyout professionals, on how AI affects their deal work. Each survey asks different people different questions, so Qaike Ltd compare figures within a survey rather than across them. One Mercer figure matters most for this post: 63% of managers use vendors' off-the-shelf AI tools, and only 9% run a fully proprietary AI platform. Most firms will not build their own. Picking the right vendor, and checking its claims, is now the core decision. Every named win in its period sits in research, sourcing, diligence, reporting, operations or data platforms. None hands AI an investment call. Conclusion: what to ask before you trust a "client win" * Can I speak to a named client like Qaike Ltd? A logo slide is not a reference. * Firm-wide or one team? Ask how many seats are in active use six months after launch. * Who owns the vendor? Bank or rival shareholders may shape the roadmap and raise data questions. * Who measured the results? Ask for the baseline behind any time-saved figure. * What happens if the vendor is sold? Recent deals, such as LSEG's plan to buy Octus, show this risk is real. The next quarter will show whether more firms put their names to deals, or whether AI stays in wide but quiet use. To compare the vendors named here, see Qaike's vendor directory. This article is based on publicly available information and reflects Qaike's own analysis and opinions. It is not intended to provide professional advice, endorsement, or a definitive assessment of any vendor or product. While every effort has been made to ensure accuracy, completeness and correctness cannot be guaranteed. Created by Qaike - Powered by AI.

Unique AI launches AI teammates for financial services. 12 hours ago * Unique AI, a Swiss-founded secure AI platform for financial services, launched Remy, an AI teammate for financial advisors, at its New York launch event on October 6. * Remy builds a continuously updated "living memory" of each client from CRM, core banking and custody platforms, Outlook, Teams, calendars and meetings, with every fact linked to its source and date. * Each morning, Remy delivers a prioritized dashboard of recommendations and open actions, briefs for every meeting in the next week, and prepared client emails ready to send. * Chief Product Officer Dana Ritter also previewed Casey, a compliance teammate for Source of Wealth narratives, KYC questionnaires and account remediation, and Ivy, an investor relations teammate that drafts RFP and DDQ responses. * Remy, Casey and Ivy share the same living memory, so what one teammate learns about a client is available to the others. * The platform deploys in each firm's own environment, works with the firm's choice of language model, mirrors existing access permissions, and includes guardrails against mixing information between clients. * Founded in 2021 in Zurich, Unique AI is used by more than 40 financial institutions and 50,000 professionals, and the launch marks the next step in its US expansion.