KPMG Builds an AI Startup Engine Inside the Big Four

NEW YORK — KPMG LLP is dismantling traditional committee layers to create a specialized unit that will build and spin off artificial intelligence startups, as the century-old business model of the global professional services industry faces an unprecedented existential challenge. That model relies on vast hierarchies of junior associates billing clients by the hour.

Named Client Technology & Innovation (CT&I), the new group takes effect Sept. 22 and combines KPMG’s artificial intelligence, innovation, and alliance ecosystem functions in one division. KPMG has appointed Todd Lohr, a 15-year veteran of the firm, as Vice Chair of CT&I. He will report directly to Chief Executive Officer Tim Walsh, an unusual arrangement for the tradition-bound “Big Four” accounting firms, which also include Deloitte, EY, and PwC.

Lohr, 49, previously worked as a business architect at Zurich Insurance and was the third employee at Confiance, a financial services consulting startup he co-founded during the 2008 financial crisis. At KPMG, he established the firm’s early data and AI practices, directed its U.S. technology and cybersecurity business, managed third-party technology alliances, and led the clients-and-markets commercial organization.

The direct reporting line moves AI and innovation strategy out of the middle-management layers of functional heads and vice chairs and places it on the chief executive’s immediate agenda. The restructuring coincides with the retirement of Steve Chase on Sept. 30. Chase built KPMG’s initial AI and Digital Innovation group in 2023 and most recently served as Global Head of AI, a position he assumed in August 2025.

For decades, elite accounting and consulting firms have used “leverage”: recruiting large classes of university graduates to conduct manual research, data analysis, and document drafting, then billing clients for their time. Generative AI threatens to automate those baseline tasks, pushing firms toward value-based pricing, proprietary software licensing, and subscription-style service models.

CT&I will address that shift through four areas: Products and Platforms Strategy, which will move beyond internal productivity tools toward commercial software and platforms for clients; Commercial Infrastructure, which will create new go-to-market structures and financial agreements beyond hourly billing; Firmwide AI and Data Strategy, which will direct the technological architecture previously handled by the AI & Digital Innovation group; and Innovation and Incubation, which will operate an internal venture studio using flexible capital structures.

Lohr’s startup background influenced CT&I’s design, which he describes as modeled directly on Silicon Valley venture incubators. The unit will incubate “edge disruption plays” that could eventually cannibalize KPMG’s core business lines. Depending on market traction, those businesses may return to KPMG’s core operations, operate as standalone joint ventures with major technology partners, or be spun out using outside venture capital. Lohr said his ultimate goal is to nurture a billion-dollar standalone company through the incubator model.

KPMG plans to rotate rising employees temporarily out of tax, audit, and advisory roles so they can work as founders and operators in the startup portfolio. When they return to core service lines, firm leadership expects them to be more tech-literate, following the classic Big Four model of talent development and alumni networks.

KPMG has spent several years preparing its board and management committee for the structural pivot by educating leaders about technological disruption. Regular leadership summits in Silicon Valley have included closed-door meetings with frontier AI research labs, chief scientists, and chief information security officers focused on the trajectory of foundational models.

The new unit will also manage KPMG’s commercial alliances with Anthropic, Microsoft, OpenAI, Google Cloud, and Databricks under a single ecosystem mandate.

The restructuring arrives as professional services firms deploy artificial intelligence while confronting serious reputational risks. In June 2026, GPTZero, an AI-detection platform founded by Princeton researcher Edward Tian, reported that KPMG International’s flagship thought leadership report, “Total Experience: Redefining Excellence in the Age of Agentic AI,” contained widespread fabrications. An investigation found that 40 of its 45 citations of real-world case studies involving organizations including UBS, the U.K.’s National Health Service, and Swiss Federal Railways did not exist. Roughly half of the report’s factual claims were either false or misattributed.

GPTZero later reported similar AI-generated hallucinations across the Big Four. At EY, 60% of the citations in a December 2025 cybersecurity report were found to be hallucinated. At Deloitte Australia, a University of Sydney academic identified 19 separate AI hallucinations in one published report. At PwC, GPTZero analysis indicated an 84% probability that a prominent Middle East governance report was entirely machine-generated.

KPMG recently deployed an internal AI auditing tool called “Hawk,” designed as an in-house equivalent to GPTZero. It scans KPMG-generated content for potential AI hallucinations before publication. A spokesperson said Hawk is currently used primarily by the firm’s marketing and thought leadership teams, while CT&I’s wider remit will include building similar quality-assurance and trust frameworks into the client-facing software products developed by its edge startups.

Lohr helped launch KPMG’s initial “Trusted AI” advisory offering nearly a decade ago. He has emphasized that technological tools cannot replace professional accountability and, while acknowledging generative AI’s transformative potential, reiterated that individual professionals must remain ultimately responsible for the integrity of their work product.