Every few weeks, a CEO or CHRO asks me some version of the same question: “What’s the right way to organize around AI?”
They usually want a blueprint, a model to copy, a shortcut to certainty. The problem is, the playbook doesn’t exist. And by the time Bain or McKinsey publishes theirs, it’ll already be obsolete.
What makes the question so persistent is that AI exposes uncomfortable truths about how companies actually operate. It forces decisions about who controls data, who absorbs risk, and who gets to say no when experimentation collides with regulation, legacy systems, or quarterly pressure. An org chart feels like a way to resolve those tensions cleanly, even though it rarely does.
So CEOs are improvising. Some are putting AI under the CDIO. Others are hiring Chief AI Officers. A few are creating entirely new “AI business units.” The debate isn’t just academic; it’s shaping billions in investment and influencing who gets a seat at the leadership table.
And few have sparked more of that debate than Jamie Dimon.
When Jamie Dimon Speaks, Boards Listen
When Dimon said that AI should report directly to the CEO, it became an instant boardroom soundbite, a simple answer to a complex question. To his credit, he’s right about one thing: AI is a business transformation, not a technical initiative. It belongs at the center of the strategic conversation, not buried three layers down in IT.
But here’s where things get fuzzy.
If you take Dimon’s statement at face value, that AI and data should “sit outside Technology,” it implies that the CIO or CTO isn’t equipped to lead AI. That may make sense for JPMorgan, a $500 billion institution with 55,000 technologists and a world-class data foundation. For most companies, though, that separation would break the connection between innovation and execution.
The truth is that JPMorgan’s model works not because AI reports to the CEO, but because the company has mastered the three capabilities that make AI real, the same ones most companies struggle to synchronize.
I’ve seen the opposite play out repeatedly. A company announces a Chief AI Officer with CEO visibility but no control over data platforms, engineering capacity, or operating budgets. Six months later, the role owns a roadmap but not delivery, while the CIO owns delivery but not prioritization. Pilots proliferate, production stalls, and the board concludes the technology was overhyped when the real issue was structural misalignment.
The Three Core Capabilities of AI Leadership
Every company chasing AI maturity needs three muscles that move together: Explore, Scale, and Embed.
Each serves a distinct purpose, and the magic happens when they operate in concert.
- Explore is where curiosity meets direction. It’s the work of discovering what’s possible: testing, experimenting, and translating emerging technology into strategic opportunity. The best organizations explore with purpose: they don’t chase hype; they scan for signal.
- Scale is where ideas meet infrastructure. It’s the muscle that builds, the data platforms, systems, and tooling that take innovation from pilot to production. It’s about creating reliability, repeatability, and speed without losing control.
- Embed is where change becomes impact. It’s the human work, aligning incentives, governance, and behaviors so AI becomes part of how the business operates, not just another experiment.
This is also where many AI efforts quietly fail. If incentives still reward manual decision making, if governance treats models as exceptions rather than systems, or if frontline leaders are not accountable for adoption, even technically successful deployments stall. Embedding is not culture work in the abstract. It is operational redesign.
The most effective companies don’t treat these as silos. They manage them as a living system: exploration generates ideas, scaling turns them into capability, and embedding ensures they endure.
What JPMorgan Got Right (and Why It Works for Them)
JPMorgan happens to have all three muscles firing, even if they don’t describe it that way.
- Lori Beer, Global CIO, leads 55,000 technologists focused on automation and the firm’s horizontal technology platform. She represents Scale, building the systems and infrastructure that make AI real across the enterprise.
- Teresa Heitsenrether, Chief Data & Analytics Officer, drives enterprise-wide adoption of data, analytics, and AI. More strategist than technologist, she represents Embed, ensuring new capabilities take root and create sustained value.
- Manuela Veloso, Head of AI Research, is a world-class academic exploring quantum computing and next-generation models, leading the Explore capability that keeps the company on the frontier of what’s possible.
None of them reports directly to Dimon. Beer and Heitsenrether report to the COO, with dotted lines to the CEO. Veloso sits further down the structure but has direct access when needed.
So when Dimon says AI “sits outside of Technology,” it’s more philosophy than org chart, a reflection of how integrated & mature the system already is.
That’s what most people miss when they quote him. The magic isn’t the reporting line. It’s the system of capabilities that works behind it.
Why Context Matters
For JPMorgan, separating AI and data from Technology makes sense. The tech foundation is already strong enough to sustain it. The muscle memory exists.
For everyone else, context matters. The question is not whether JPMorgan’s model is right or wrong. The question is whether your organization has earned the right to copy it.
If your technology stack isn’t modern, if your data is fragmented, or if your teams still think in projects instead of products, pulling AI “out of Technology” is like taking the engine out of the car. It won’t move.
In those cases, keeping AI within Technology, but elevating its visibility and strategic sponsorship, often makes more sense. You don’t need to mimic JPMorgan’s hierarchy. You need to build JPMorgan’s alignment.
Because what ultimately matters isn’t who AI reports to, it’s whether your organization can explore, scale, and embed in sync.
So What Should Leaders Take Away?
Jamie Dimon is right about the why. AI and data are too important to be delegated. They need the sponsorship and attention of the CEO. But the how is contextual.
The companies that will win with AI aren’t the ones with the flashiest titles or newest org charts. They’re the ones that get the choreography right, connecting strategy at the top, technology at the base, and leadership in between that can translate between the two.
In that sense, Jamie Dimon’s structure isn’t a playbook. It’s a proof point, evidence that when the right muscles move together, AI can stop being a headline and start being an advantage.
Note: JPMorgan Chase is not a client of SPMB. All insights referenced here are based on publicly available information and conversations with industry sources.
