The Trust Gap Investors Should Price In
The headline number is striking: within two years, 100% of surveyed enterprises plan to be using agentic AI, with 69% expecting to use it widely.1 But on average, those same organizations' AI systems can only reach 45% of company data — and at firms categorized as 'data laggards,' that access falls to 30% or less.1 The gap matters financially because it is a direct constraint on ROI: two-thirds of data laggards say legacy data systems limit their ability to scale AI agents (66%) and prevent agents from making decisions at the speed the business needs (68%), versus just 8% of 'data leader' organizations reporting either constraint.1
A note on how much weight to put on those figures: they come from MIT Technology Review's "Scaling AI agents with trustworthy data,"1 a source whose claims Via News's verification process has checked against source-of-truth data 11 times — and none held up as stated. That does not mean the directional story is wrong; survey-based industry reporting of this kind is difficult to fully source-trace, and the pattern it describes (fast intent, slow data readiness) is consistent with what enterprise vendors are independently building toward, as below. But readers should treat the specific percentages as indicative, not verified fact, and that distinction is itself the story: capital is being allocated against numbers the market hasn't independently checked.
Platform Players Underwrite the Access Problem With Balance-Sheet Commitments
Where the data-access constraint is real, large platform vendors are responding with long-duration enterprise contracts rather than point products. Manulife and Microsoft renewed and expanded their partnership under a five-year agreement, with Manulife adopting Microsoft's Frontier Suite, deploying Microsoft Agent 365, and expanding Microsoft 365 Copilot to more than 30,000 employees.2 Shamus Weiland framed the commitment in governance terms: "Our partnership with Microsoft is a critical enabler of Manulife's continued evolution into a truly AI-driven organization. Adopting the Microsoft Frontier Suite represents the next phase of that transformation, giving us the trusted foundation to advance AI across our global operations with confide[nce]."2 A five-year term on an enterprise AI governance contract is itself a capital-allocation signal — it is a multi-year opex commitment, not a pilot budget line.
Box moved on the same trust problem from the content-security side, announcing agent guardrails, third-party agent activity oversight, prompt injection detection, and classification-based access policies for enterprise content.3 Nomura Research Institute's Tatsutoshi Murata linked the value directly to risk management as agent usage scales: "As we rapidly advance our utilization of AI agents, we expect Box — which has consistently led the development of security management capabilities for secure collaboration — to provide the administrative features needed to safely leverage this new era of AI. In particular, we've found it extremely reass[uring]."3 (Both the Manulife and Box announcements were distributed via NewsEOD, a wire whose claims Via News has checked 2,547 times with a 33% hold-up rate — read the specific figures above as company-reported claims, not independently confirmed metrics.)
The Vertical Bet: Monetizing Narrow, Auditable Workflows
A separate and, for finance readers, more directly investable pattern is emerging among venture-backed specialists that sidestep the data-access bottleneck entirely by targeting narrow, well-bounded back-office processes. Maisa AI's CEO David Villalon defines the company's addressable market explicitly around regulated-industry process automation: "I define my market like the market of process automation of core business and production tasks at regulated industries. So, at the end, it's all the core tasks that are today being manual or handled by humans that are part of the core product or the core services of the company. So, that means back [office, operations, finance]."4 The regulated-industry framing is the point — those are precisely the workflows that must be auditable and reproducible, which is also why they are less exposed to the 45%-data-access ceiling: the task and its inputs are already narrowly defined.
Penguin AI is sizing the healthcare back-office opportunity in comparable terms. Head of Marketing Glenn Herzberg puts the addressable spend at roughly $1 trillion a year — about a quarter of total US health spend — with published estimates attributing around $570 billion of that to administrative work that has no effect on health outcomes.5 That $570 billion figure is the commercial thesis in one number: it is spend with no clinical defender, which makes it the cleanest possible target for an automation vendor pitching cost reduction rather than outcome improvement.
Casap is applying the same logic to fintech dispute resolution. Primary partner Emily Man, whose firm backed the company, described the founding team's motivation: "They had both experienced the pain points of disputes firsthand at their respective large fintech companies and saw like the amount of internal effort and organizational work that it took to solve those challenges even with a really strong engineering team."6 The founders came from Robinhood and Chime — operators who had already tried and struggled to solve the problem with in-house engineering, which is itself a data point on how hard the workflow is to automate cheaply. Man's diligence leaned on reputational signal as much as product: "the feedback was just resoundingly clear that these were two exceptional builders who were really passionate about starting their own thing and solving problems that they had seen before,"6 and Primary's initial interest was driven by the founders' backgrounds and the opportunity itself: "We were immediately really excited about them because of their backgrounds, and they talked to us about this opportunity that they were thinking about tackling."6
What to Watch
Two financial questions separate from each other going forward. First, whether the platform-partnership model (Manulife/Microsoft, Box's enterprise customers) converts multi-year governance contracts into measurable productivity gains, or whether the 45%-data-access ceiling reported by MIT Technology Review's survey1 — treated with the reliability caveat above — continues to cap realized ROI regardless of contract size. Second, whether the vertical specialists' bet holds: that dispute resolution, healthcare administration, and regulated back-office finance are narrow enough to automate profitably without solving the broader enterprise data-access problem at all. Penguin AI's $570 billion administrative-waste figure5 and Maisa's regulated-industry framing4 are the numbers to track as these companies report customer counts and funding rounds — they are the market-sizing claims the vertical thesis has to prove out.
