Notes from the Sidebar Summit at Nasdaq, where the IPO conversation finally stopped being theoretical

Friends,

I spent yesterday afternoon in New York at the Nasdaq MarketSite for the Sidebar Summit, and I moderated a panel on markets, AI, and capital formation. Sidebar Summit was part of a series of events comprising New York Tech Week. I want to share what I took away, because the timing of the day turned out to be almost too good to script.

A few hours before I walked on stage, Anthropic filed a confidential S-1 with the SEC. The company behind Claude. Worth something like nine…hundred,,,billion…dollars. Sixty…five …billion…dollars – raised in a single round the week before. The periods are illiteration to make a point. These are giant numbers, and the public statement said almost nothing. No price, no date, no share count. It might go public after the SEC finishes its review, market permitting. That is the whole message.

So I opened with it. I told the room that the most valuable private company most of them have never owned a share of (unless they participated in a pension fund) just started the most expensive game of will-they-won’t-they in history. SpaceX is already on file. OpenAI is reportedly next. Three of the largest offerings ever attempted, all lined up for the same narrow window, and not one of them willing to quote you a price. For the last three years, everyone in a room like that one has told me the IPO market is coming back. It has been coming back the way my flight out of Newark is always about to board. At some point, coming back is just a thing we say to each other.

That set up the real question for the next 45 minutes. Is what happened that morning the window finally opening, or is it three companies big enough to kick the door down because the door stopped working for everybody else? And if it is the second one, what does that tell the thousand companies behind them that cannot force anything?

The panel before us framed the stakes

Our session followed a strong panel led by Rob Gould Protiviti on whether and when to go public, with Bill Barhydt of Abra and Aaron Schumm of Vestwell. Bill and Aaron come at the question from very different businesses, but they landed in the same place I keep landing with clients: going public is no longer a milestone you reach because you are supposed to. It is a financing and liquidity decision you make with eyes open, weighed against staying private, against an M&A exit, against the structured secondaries and continuation vehicles that did not exist when I started doing this. The decision has gotten harder, not easier, and companies only do it when they have to.

What my panel actually argued about

I had told the audience there would be no slides and no prepared remarks, and that my panelists disagree with each other on things that matter. They delivered. I will not pretend every exchange resolved cleanly, because it did not, and that was the point.

The two lines I am still thinking about both came from the stage.

The first was from Mark Gorenberg, chairman of the MIT board of trustees and founder of Zetta Venture Partners. When the conversation threatened to drift into all the reasons it is hard right now, the cost of being public, the asymmetry of private capital, the policy gridlock, Mark cut through it. His message was simple and he delivered it with the conviction of someone who has watched several technology cycles crest and break. It is time to get going. This is the greatest era we have ever been in, and the best time in the history of the world to form a new technology company. I have heard a lot of market commentary in 25 years of doing this work. That one stuck, because it reframes the entire anxiety of the moment. The structural problems in the public markets are real. They are also not a reason to sit on your hands while the most consequential platform shift of our lifetimes plays out.

The second came from Chris Kelly, and it was advice for founders rather than market color. Test out every new idea. In an era where the cost of building has collapsed and the tools to prototype are sitting on every desk, the constraint is no longer capital or even engineering time. The constraint is willingness to try. The companies that win the next decade will be the ones that ran the most experiments, not the ones that waited for certainty that was never coming.

Put those two together and you have the whole thesis of the afternoon. Get going, and try everything.

The fireside that followed, and a client I am proud of

After our panel, Gerald Brady of SVB sat down with Deven Parekh of Insight Partners for a fireside chat, and I will admit I enjoyed it for a slightly self-interested reason. Insight backed our client Mirantis, and I got to brag a little. Mirantis spent more than a decade as a cloud infrastructure company, then executed a genuine pivot into AI, and last month sold to IREN in an all-stock deal valued at roughly $625 million in stock as of signing. IREN is using the acquisition to climb the stack from raw GPU capacity into full AI cloud delivery, and Mirantis brought the orchestration software, the enterprise customer base, and the operational depth to make that credible. It is exactly the kind of outcome this market is supposed to produce and too rarely does: a company that read where the puck was going, made a hard pivot, and found the right home at the right time. The Insight team helped get them there.

Which brings me to the part I could not stop talking about

I came into the day a structural skeptic about the IPO machine, and I left one. Filing is not pricing. A confidential draft starts a clock and commits to nothing. The gap between the three giants tiptoeing toward the public market and the founders sitting in that audience wondering when their own window opens is the gap that policy still has not fixed. And it is worth being honest about why that gap exists, because the answer is not the one most people reach for.

The public market has been shrinking for thirty years. The number of U.S.-listed companies has fallen from roughly 8,000 in the mid-1990s to about 4,000 today. The capital did not vanish. It moved to the private markets, which now sit on something like $8.5 trillion and ask for almost none of the disclosure or accountability the public markets require. A generation of growth has happened in a room most Americans are not allowed to enter. The teacher and the firefighter saving through a 401(k) get to buy in late, after the best gains have already been distributed to a narrow set of insiders. That is the real cost of the shrinking public square.

SEC Chairman Paul Atkins has made reversing the trend the center of his agenda, under the banner “Make IPOs Great Again,” and in a single month this spring, the SEC turned that slogan into actual rulemaking. Semiannual reporting as an option. Shelf registration opened to nearly every company instead of just the giants. A simpler filer framework, relief from the most expensive recurring audit ritual, a multi-year on-ramp for the newly public, and a withdrawal of the climate-disclosure rules a federal court had already paused. I support all of it. Then on May 26, the Chairman went to Stanford and opened a comment file asking for bold and creative ideas to modernize how companies go public. That invitation matters as much as the rules.

But here is the uncomfortable truth, and it is the thing I kept pressing on stage. Companies do not avoid the public markets mainly because reporting is expensive. They avoid them because the private markets now offer everything a growing company needs with none of the friction. You cannot draw companies back by discounting the rent when the alternative is unlimited and unregulated. Lowering the cost of being public is half the job. The other half is removing the barriers that keep companies from going public in the first place, and on that, the work has barely begun.

So when people ask me what I would actually add to the agenda, here is my list.

• Rewrite the gun-jumping rules. The rules governing what a company can say while going public were written in 1933 and last meaningfully touched in 2005. They force a company into silence at the exact moment investors most want to hear from it, and they turn ordinary communication into a minefield only expensive counsel can cross. Rebuild the whole thing around one principle: police fraud, not the act of speaking.
• Revive research and trading for smaller companies. Wall Street has almost no economic reason left to publish research on, or make markets in, anything outside the largest names. You can see the result in the concentration of the market, where seven companies made up roughly a third of the S&P 500 at the end of last year. A company that goes public only to find no analyst covering it and no real market in its shares has little reason to be public at all. Let smaller issuers choose wider trading increments, and support the firms willing to commit capital to make markets in their stock.
• Fix the litigation calculus. A company doing an IPO is exposed to a class of securities suits that seasoned public companies are largely shielded from, and the safe harbor that lets mature companies discuss their projections in good faith does not extend to the companies that need it most. Until honest forecasting is something a newly public company can do without inviting a strike suit, the math will keep pushing companies private.
• Modernize the path to listing. After the Supreme Court’s 2023 decision in Slack Technologies v. Pirani, the heavy registration statement once required for a direct listing adds cost without adding much protection. Streamline it. And rehabilitate, rather than bury, the SPAC and the alternative public offering. Both are legitimate routes to the public market that the current rules treat as second-class.
• Make the giants disclose. There is no good reason a company that has raised a billion dollars or more in private capital should sit outside the public reporting system. The money funding those companies comes from pension funds, endowments, insurers, and the vehicles managing ordinary Americans’ retirement savings, which are exactly the investors the securities laws exist to protect. A company of that size is public in substance. It should disclose like one, on terms fitted to its circumstances rather than copied wholesale from the rules for a Fortune 100 issuer. I will let you decide which company that filed a confidential draft this morning might have something to say about that.
• Welcome the world. America’s share of global IPOs has fallen sharply over two decades. An agenda serious about capital formation cannot treat foreign issuers as an afterthought. We should be competing to list the world’s growing companies here, not narrowing the door.

The takeaway

Put it together and you get the through-line of the whole afternoon. Mark is right that this is the best time in history to build, and that the only unforgivable move is to wait. Chris is right that you test every idea, because the cost of trying has never been lower. And the policy job is to make sure that when those companies are ready, there is a public market worth going to, one that ordinary Americans can actually invest in.

The May agenda is the most serious attempt in a generation to reopen the U.S. public markets, and it deserves broad support. But finishing the job means fixing the rules that keep companies out, not just discounting the rent for the ones already inside. The comment windows close this summer. If you want a deeper, more open public market, this is the moment to say so on the record.

Get going. Test everything. And help us finish the job.

More soon.

Thanks to Vishal Verma and the Sidebar Summit team for producing a great event.

Louis Lehot is a partner at Â鶹´«Ă˝ based in Silicon Valley, where he advises entrepreneurs, investors, and public companies on corporate, securities, capital markets, and M&A matters. He has more than 25 years of experience guiding companies from formation to IPO and beyond.

Opinions are the author’s own and not those of his firm. This is commentary, not legal advice. Attorney advertising. Prior results do not guarantee a similar outcome.

I have spent close to three decades sitting across deal tables from private equity sponsors, and for most of that time the game looked the same. Buy a good business, fix the capital structure, drive operational discipline, dress it up, and sell it for a multiple of what you paid. Financial engineering on one side of the table, operational engineering on the other. That was the playbook, and it worked.

It still works, but it is not enough on its own anymore.

What I am seeing in deal rooms across Silicon Valley and the rest of the state is that sponsors have stopped opening with the old questions. The first question I get now is some version of, where does AI live inside this business, and what is it worth if we get it right?

Let me put some context around that, because it is not theoretical. It is a function of where the market actually sits right now.

The exit door is jammed

Coming into 2026, the industry is living with the consequences of three years of stretched hold periods. DPI, distributions to paid-in capital, has become the metric every LP wants to talk about and that very few GPs want to discuss. The prolonged government shutdown at the end of 2025 took what should have been a strong Q4 IPO window and turned it a lost opportunity. Companies didn’t disappear. They stacked up. Figma cleared the window. CoreWeave cleared the window. Cerebas cleared the window.  The rest are queued, and they are going to stay queued until interest rates come down and risk-on sentiment returns to support the multiples we need to clear the backlog.

When you cannot get out the front door, you find other ways. We saw a real wave of secondaries last year, with sponsors packaging assets into continuation vehicles, sometimes single asset, sometimes baskets, and selling them into other funds. I worked on more of those in 2025 than I had in any twelve-month period in my career. That tool works once or twice, but it does not solve the underlying problem, which is that you have to actually create value during the hold. And in a slower exit environment, you have to create more of it than the model said you needed when you signed the LOI.

That is what is really driving the AI conversation in PE right now. It is not about chasing a trend. It is about defending and expanding margin during a hold that is lasting longer than anyone underwrote.

AI has moved from experiment to operating budget

Across my portfolio company clients, AI has stopped being a slide in the strategic plan and started being a line item. Back office automation, forecasting, supply chain optimization, customer service deflection, code generation inside engineering. I think the lean manufacturing crowd from the 1990s would have a hard time recognizing what is happening to SG&A inside well-run portfolio companies right now.

I had a conversation recently with an operating partner at a mid-market sponsor who told me about their thesis that every new platform deal now starts with one question. What percentage of headcount is doing repetitive cognitive work that an agent could do tomorrow? Two years ago, that was a footnote. Today it is the model.

The numbers track with what I am seeing on the ground. AI and machine learning private equity deal value went from roughly $42 billion in 2023 to north of $140 billion in 2024, and the momentum kept building through 2025. Q1 of 2026 broke every record on the books, with AI taking roughly 80 percent of the venture dollars deployed. When capital moves that decisively, sponsors who are not running an AI-first value creation thesis are going to find themselves selling into a market that prices one in, whether they delivered it or not.

Robotics is no longer just industrial

I came up in the Valley watching robotics live in factories and warehouses. That is not where the next leg is. The advances in computer vision, autonomous systems, and sensor technology have opened up applications that just were not economically viable five years ago. Logistics yards, food service, construction, field services, healthcare delivery. Defense tech and space are particularly interesting right now, because when you sign a government contract you have visibility on five to ten years of revenue, and that kind of stability is hard to come by in AI infrastructure, where people in PE still cannot really bank on what the next five years look like.

For sponsors looking at industries with persistent labor shortages, wage inflation, or seasonal volatility, robotics is a credible operational lever, and it is the kind of investment that can change a business’s unit economics permanently. Permanent margin expansion is exactly what you need when you may be holding the asset another two or three years longer than your model contemplated.

A diligence war story

The diligence I am running on technology-enabled deals these days does not look anything like the diligence I was running in 2022.

We spend real time now on data rights. Whose data is the model trained on, do you have the license, was it scraped, are there indemnities, and are the outputs clean for commercial use? On a recent platform acquisition, we found the target had been training an internal model on a dataset that, when we peeled the onion, included licensed third-party content with field-of-use restrictions that flatly prohibited model training. The deal did not die, but it got repapered, and the seller took a haircut. Founders should assume that their data rights are now getting the same scrutiny in diligence as the cap table.

We spend real time on regulatory exposure. The EU AI Act is being implemented, and state laws on automated decisioning are proliferating in California, Colorado, and New York. If a portfolio company is using AI to make hiring, pricing, credit, or insurance decisions, there is a regulatory diligence problem there, whether the sponsor knows it or not. I am also seeing reps and warranties insurance carriers begin to underwrite AI-specific exclusions, which tells you the market is pricing this risk in real time.

We spend real time on cybersecurity, because the attack surface of an AI-enabled enterprise looks nothing like the attack surface of the same business pre-AI. Prompt injection, model extraction, training data poisoning. I am taking incident response calls on this stuff right now.

And we spend real time on workforce. Mass automation has labor consequences. WARN Act considerations, plant closing rules, severance practices, and in unionized environments, bargaining obligations. A sponsor who plans to take headcount out of a target needs a plan that survives legal scrutiny, not a slide deck.

A tale of two worlds in the talent market

At the TED AI conference last year, I described what I have been calling a tale of two worlds in AI. The hyperscalers and the large model labs are paying compensation packages that early-stage startups simply cannot match, and I have watched a Series A founder lose his lead engineer to a nine-figure package from one of the frontier labs. That is the market we are operating in.

For PE sponsors, this matters in two ways. When you buy a tech-enabled platform, your retention plan for technical talent has to be designed for a world where Google, Meta, and the AI labs are calling your engineers every Tuesday. Single-trigger acceleration, aggressive equity refresh grants, and meaningful cash retention, which used to be reserved for the founder and the CTO, are now what it takes to keep the top of the engineering org in place. And when you are diligencing an AI-native target, founder-engineer concentration risk is real in a way it has never been. Lose two people and you may not have a company.

Operational engineering has become technology engineering

The sponsors who are going to outperform in this cycle are the ones who function less like financial buyers and more like technology transformation platforms. They have operating partners with real AI implementation experience, they have relationships with the model providers and the systems integrators and the automation vendors, and they walk into a target and know within thirty days which workflows to automate, where to deploy agentic systems, and where to pull back.

Sponsors who treat AI as a cost-cutting tactic will get the cost-cutting return, which is a one-time bump that gets competed away inside eighteen months. Sponsors who treat it as a strategic platform investment will get something more durable. Margin expansion that sticks, scalability without proportional headcount, and an exit story that resonates with the next buyer in a market that, by then, will be pricing AI delivery into every model.

Where this leaves us

I tell my PE clients the same thing I told a roomful of them last quarter. Hope is not a strategy, preparation is not optional, and the best deals are won long before the term sheet arrives. That has always been true. What is new is the playing field. AI, automation, and robotics are not a side dish to value creation in this cycle. In a lot of cases, they are the value creation, and the sponsors who treat them that way are the ones I would put my money behind.

The convergence of private equity, artificial intelligence, and robotics in Silicon Valley and across the rest of the country is real, and it is accelerating. Sponsors who internalize that, who build their teams around it, structure deals around it, and put it at the center of diligence, are going to win this cycle. Those who do not are going to spend a lot of time explaining to their LPs why DPI never came back.

In 20-plus years of papering venture and private equity deals in Silicon Valley, I’ve sat across the table from a lot of CFOs. The good ones tell you what’s actually moving the business in the first ninety seconds. The great ones tell you, in the same breath, what they’d change about the last cycle. On Thursday evening at the Wells Fargo Innovation Center in Menlo Park, I had the privilege of moderating a closed-door conversation with four of the great ones, and it was the most useful ninety minutes I’ve spent on the state of value creation in 2026.

The credit for getting that room in the room belongs to the organizers. Christina Bui at Robert Half built the panel, ran the logistics, and kept the program on rails. Armello Rodriguez and the Wells Fargo Innovation Center team hosted us and made the venue feel like a working room rather than a stage.  You can see it in the photo above, panelists on stools, no podium, the program board behind us setting the week’s tempo. And Murray Newlands, who knows everyone worth knowing on the early-stage capital side, helped pull the right people into the seats and then sharpened the questions in the lightning round.

The panel itself was deliberately tight. Celine Dinh of Brain-Life and Vietsol. Drew Hamer of Arcserve. Ted Marks of Applitools. Viraj Patel of Signeasy. Four PE-backed CFOs who had pulled real levers in the last twelve months. No theory, no slideware, no AI-washing tolerated. Here’s what came out of it.

The macro: AI is on fire, SaaS is in a squeeze

I opened the night with a framing the room didn’t push back on. AI is in a boom cycle:  budgets are flooding in, board attention is everywhere, infrastructure spend is accelerating.  Meanwhile, traditional SaaS is in a tougher cycle than it’s been in a decade. Growth multiples have compressed. Expansion is harder. Churn is scrutinized more than ever. The math of the Rule of 40 has become the Rule of 100.

For a PE-backed CFO, that creates a very specific mandate: create value now (in EBITDA and in cash), protect the downside, and keep optionality for an exit that may take longer than anyone wants. Or, the way I kept saying it through the night: value creation must be manufactured, not assumed.

The ground is moving on valuation

The most uncomfortable truth in the room (and the one every CFO nodded at) is that the ground has shifted beneath our feet on how growth gets valued. Thirty percent growth used to clear the bar without a second look. In 2026, the same number provokes a different question: at what margin, with what retention, and at what efficiency?

Murray pushed it in the lightning round: is 30% growth respectable, or irrelevant? The stance back was honest.  It depends entirely on whether the growth is efficient and durable. A growth number without a margin profile and a retention story is no longer a complete sentence. From where I sit on the deal side, that’s exactly the shift I’m seeing in term sheets and quality-of-earnings work. Buyers and sponsors aren’t paying for top-line theater. They’re paying for predictable, repeatable margin expansion at credible scale.

That reframes everything a CFO does. It pushes pricing discipline higher up the priority list. It makes churn analysis a weekly ritual, not a quarterly autopsy. And it puts an enormous premium on what Drew called margin that sticks, the structural improvement that doesn’t reverse the moment you stop cutting.

EBITDA from precision, not pain

Every CFO in the room had already cut what was easy to cut. The interesting question was what comes next. The answer, almost unanimously: pricing, mix and operating discipline.

Viraj walked through what tightening pricing discipline actually looked like:  fewer discounts, clearer packaging, and enforcement in the approval workflow, so the policy actually held. The result was contribution margin expansion without slowing top-line growth materially. That’s the move every sponsor wants to see, and frankly the one that survives a buyer’s diligence: margin you can defend, not margin you had to bleed for.

Ted raised the warning flag too. “EBITDA at all costs” backfires. CFOs misjudge the tradeoff when they cut into revenue retention or product quality.  The churn shows up two or three quarters later, and by then you’ve also lost the flywheel. I’ve seen this exact pattern surface in working-capital adjustments and earn-out disputes more times than I can count. Sponsors reward repeatable margin, not one-time austerity.

The cultural shift: finance has moved into the operating core

The cultural shift inside finance departments is real, and it’s bigger than tooling. Celine described moving from monthly finance reviews to weekly operating reviews tied to a handful of leading indicators.  Crucially, finance owned the scoreboard and the decision gates, not just the report-out. Pricing exceptions, hiring approvals, spend governance: those gates moved into finance. The operating teams own the actions; finance owns the discipline and the follow-through.

Drew delineated the boundary line cleanly: finance should lead where the decision is economic, and support where the decision is technical or customer-facing. The friction shows up when finance tries to own the how instead of the what and the economics. That’s a useful guardrail.  It’s how a CFO becomes a strategic operator rather than a senior accountant with a bigger title. Increasingly, it’s also how a CFO becomes a credible co-pilot to the sponsor on portfolio company decisions.

AI, AI-washing, and the pricing question

Then we got to the segment everyone came for: AI as a margin driver, and the AI-washing problem.

Here’s the honest summary. AI is real, but it’s real in narrow workflows. Drew described automating high-volume back-office processes, requiring a payback plan, and tracking it monthly like any other capex. That works. Generic AI bolt-ons that don’t change workflow behavior, on the other hand, don’t get adopted. Viraj was blunt about that. If it doesn’t reduce time-to-value or improve retention, it’s a feature, not a lever.

Ted offered the cleanest capital-discipline filter I’ve heard: clear owner, clear KPI, defined payback horizon, and non-negotiable governance on data policy and risk. If it’s not tied to margin, retention, or cash conversion, it gets deferred. Full stop.

Then came the part I think will define the next twelve months: pricing AI services is a whole new ball game. The old SaaS playbook (based on per-seat, predictable, expand-with-headcount metrics) doesn’t map cleanly onto AI products where cost-to-serve scales with usage and value scales with outcomes. CFOs are quietly rewriting their packaging from scratch. Hybrid models, outcome-linked pricing, consumption tiers with guardrails to prevent gross margin from collapsing under heavy users.  None of this is settled. From the deal-lawyer chair, I can tell you that revenue-recognition policies, commercial contract templates, and MSAs are going to need to be rewritten right alongside the pricing pages. The companies that figure it out first will set the curve for everyone else.

This is also where the AI-washing problem bites hardest. A logo on a slide doesn’t change a P&L, and it certainly doesn’t survive a buyer’s quality-of-earnings review. Adoption metrics, cycle-time deltas, retention impact — those do. The CFOs in the room were running their own internal AI pilots the same way a sponsor runs a portfolio company: stage gates, KPIs, kill criteria. Theater gets cut early. Workflows that actually move a number get more capital.

Exit readiness when the exit is further away

We closed on optionality. If exits take longer, the winners are the teams that stay decision-ready. Ted listed the non-negotiables:  forecast credibility, clean KPIs with one source of truth, close speed and controls, working capital and cash discipline. As anyone who’s worked an M&A process knows, you can’t sell a business that can’t report on itself, and buyers pay for predictability.

Drew pushed on what “investor-grade” actually means in practice: a tighter close with standard reconciliations, a KPI pack that ties to the financials and stays consistent month to month, controls that prevent surprises rather than just detect them. Celine framed the investment question well: fund what compounds, cut what flatters.

The takeaway

Three things I’m carrying with me from Thursday night. EBITDA expansion in 2026 is coming from precision — pricing, mix, operating discipline — not from cutting. The CFO role in PE-backed companies has moved into the operating core, owning the scoreboard and the gates. And in a slower exit market, optionality is something you build early, not something you summon when the window opens.

One message to take back, in lawyer’s terms and CFO’s terms alike: be decision-ready before the market is.

The networking that came after the program was, frankly, the other half of the value. The people in this photo — panelists, organizers, sponsors, and a few friends of the program — are the kind of working network that helps a PE-backed CFO get a real answer to a hard question on a Tuesday morning, not a LinkedIn opinion. That’s what closed-door rooms are for.

Thanks again to Christina, Armello, and Murray for putting it together — and to Celine, Drew, Ted, and Viraj for being generous with what’s actually working.

AI adoption across industries is entering a new chapter. As generative AI becomes embedded in business operations, a more powerful and more complex wave is emerging: agentic AI. These systems do not merely  assist with discrete tasks, but they can autonomously execute multi-step processes, make decisions, and interact with enterprise tools across functions and sectors. For organizations navigating the shift raises a fundamentally different set of questions around governance, security, accountability, and operational resilience. The latest installment of our GC lunch series brought together legal and cybersecurity leaders from Sophos, Securonix, and RingCentral to examine how companies are preparing for this agentic era and the cross-sector challenges it presents. Here are the key takeaways.

What Are Agents, and Why Should GCs Care?

• Agentic AI is qualitatively different from a chatbot or a document review tool. Agents can take action by reasoning, planning, executing multi-step processes, accessing enterprise systems, and making decisions with varying degrees of autonomy. That distinction should get every GC’s attention fast.
• A conventional enterprise AI system might draft a contract clause when prompted. An agentic system might identify the need for a clause, draft it, route it for approval, and update a contract management system without a human initiating each step. The governance implications of that leap are profound.
• From a GC’s perspective, the defining characteristic of agents is autonomy. Autonomy, in a legal context, is inseparable from questions of authority, liability, and control. The law of agency may get tested in new and unexpected ways.

Where Agents Are in the Ecosystem

• The push to build, develop, and deploy agents is accelerating across multiple verticals, business units, and sectors. Organizations represented at the lunch series spanning cybersecurity platforms and cloud communications, are actively building and deploying agents within their product suites and internal operations.
• The maturity curve for agents is still early, but the trajectory is steep. The biggest technical developments driving the move toward agentic systems are happening now, and panelists expect agents to show up most prominently across the enterprise in the next twelve to twenty-four months.
• The ecosystem is not settling on a single model. The conversation is moving between single-purpose agents, multi-agent systems, and orchestration layers. GC’s need to understand these architectural choices because each carries different risk profiles.
• Cybersecurity companies like Sophos and Securonix are playing a dual role: both shaping the agentic ecosystem as builders and serving as frontline defenders against the risks agents introduce. The message from the panel was clear – separate the hype from what is durable, and make sure your legal team is tracking the developments that most companies probably aren’t.

Agents Are Creating a New Governance Conversation

• The global legal landscape around AI is fluid, and agents are creating governance challenges that are materially different than those posed by standalone generative AI systems. GC’s cannot simply extend existing AI policies to cover agentic deployments because the risk profile is different in kind, not just degree.
• Autonomy and Accountability. Organizations must define the scope of authority granted to an agent with the same rigor they apply to human delegation. “Least privilege” takes on new meaning in the context of agents, and decision rights need to be documented when agents are used in critical workflows. The harder question of how much discretion an agent should have before mandatory human escalation is required has no one-size-fits-all answer, but it demands a deliberate framework.
• Security. Agents introduce security risks that are unique or heightened compared to non-agentic systems. Prompt injection, tool abuse, privilege escalation, and memory poisoning all take on new dimensions in an agentic environment. Agents are, in effect, creating a new attack surface inside the enterprise and CISOs are thinking carefully about how to monitor and contain these risks. General counsel should be part of that conversation, not waiting for an incident to force it.
• Data Privacy and Additional Risks. Agents complicate existing data governance and privacy frameworks in ways that many organizations have not yet anticipated. For companies navigating the complicated global patchwork of data privacy laws, deploying agents without a clear understanding of how they interact with personal data, cross-border transfers, and regulatory obligations can present significant challenges.  
• Third-Party Risk. Contracting with agentic vendors requires a different lens. Indemnity, limitation of liability, and security provisions all need to evolve for agentic AI.

What GCs Should Take Back to Their Teams

• The panelists were candid about what excites them about the agentic future but equally candid about the governance it demands. The right balance between innovation and governance must remain dynamic, requiring an ongoing negotiation that requires engagement from legal leadership.
• GCs need to be positioned to answer a key question – if you had to explain an agent’s conduct to a regulator, a court, or your board, what documentation would you need and how much of it does your organization actually have today?
• The one thing every legal leader in the room was urged to take back to their teams and their boards was that securing the agentic future is not a technology challenge alone. It is a governance, risk management, and leadership challenge as well.  

Agentic AI is moving from concept to deployment faster than most legal departments are prepared for. The GC’s who pair early engagement with deliberate governance frameworks and who define authority, demand auditability, and insist on security by design will be the ones who lead their organizations through this transition rather than react to it.

for a timely Wall Street Journal read echoing many of the points we addressed. Onwards!

Three Key Takeaways

• AI joint ventures are creating internal portfolio conflicts for private equity firms by accelerating disruption of legacy SaaS investments they still own.
• Delaware’s SB 21 gives sponsors more procedural certainty for affiliated transactions, but only if firms implement strong governance safeguards, independent approvals, and documented conflict-management processes.
• Antitrust scrutiny of AI investments is intensifying, with regulators increasingly focused on overlapping board seats, cross-holdings, information-sharing rights, and HSR disclosures tied to AI-related deals.

This week’s headlines are remarkable. OpenAI closed a roughly $10 billion joint venture, called DeployCo, with TPG, Brookfield, Bain, and others to push its tools across sponsor portfolios. Within minutes of that news, Anthropic announced a $1.5 billion venture with Blackstone, Hellman & Friedman, and Goldman Sachs to embed Claude into mid-market businesses, starting with their own portfolio companies.

These are not passive bets. They are JVs with real governance rights that give the largest pools of buyout capital first-look operational access to frontier AI, and give the labs a distribution channel into thousands of businesses in return. The commercial logic makes sense. The legal complications are bigger than most sponsors are admitting

The Strategy Eats Its Own Portfolio

Many of the same sponsors funding DeployCo and the Anthropic venture still own legacy enterprise software businesses. Those companies were underwritten on assumptions about sticky seats, durable pricing, and net retention above 110% that AI-native services are now eroding. The tools being deployed through the new JVs will compress pricing power and shrink the addressable market of the same sponsors’ SaaS holdings. The conflict lives inside one fund family, and sometimes inside one fund.

Fiduciary Conflicts After Delaware SB 21

When a GP-affiliated AI vehicle gains value at the expense of a SaaS portfolio company held in another fund, directors of the affected entity still owe Delaware fiduciary duties to its stockholders, including LP co-investors. The legal landscape changed in March 2025, when Governor Meyer signed Senate Bill 21, amending DGCL § 144. For controlling-stockholder transactions other than going-private deals, SB 21 now provides a statutory safe harbor against both equitable relief and damages if the transaction is approved by an independent committee of at least two disinterested directors, or by an uncoerced majority-of-the-minority vote of disinterested stockholders.

Section 8 of the Clayton Act Already Sees It

The Department of Justice Antitrust Division and the FTC have unwound or prevented more than two dozen interlocks since 2022. Reporting in April 2024 indicated DOJ is specifically examining shared board service among rival AI companies.

HSR Filings Surface the Cross-Holdings

The revised HSR rules took effect February 10, 2025, and they were drafted with private equity in mind. Limited Partners with management rights now have to be named, and minority holdings in overlapping NAICS codes get specifically flagged.

Getting the JV Documents Right

Most of the legal architecture for these AI joint ventures lives inside the JV agreements themselves. Governance, information rights, valuation methodologies, liquidity rights, and transfer restrictions must all be carefully drafted to avoid future disputes and unintended control triggers.

M&A Documents Are Catching Up

AI-specific representations and warranties are rapidly becoming standard in acquisition agreements, including provisions addressing training-data provenance, open-source compliance, model dependencies, and AI risk management frameworks.

The Bottom Line

Private equity firms are racing to secure an operational edge through AI partnerships, but regulators and courts are moving just as quickly to examine the governance, antitrust, and fiduciary implications. Firms that proactively address conflicts, cross-holdings, and AI-specific legal risks will be better positioned to capitalize on the opportunity while avoiding enforcement exposure.

Key Takeaways:

• Crunchbase data shows Latin American startups raised a combined $1.03 billion in VC funding in Q1.
• Total funding was driven in large part by late-stage and growth deals, with $761 million in total.
• Mexican startups led the way with $404 million in funding, and the country boasted the largest financing round in Latin America for the quarter at $300 million.

In the first quarter of 2026, Venture Capital (VC) funding in Latin America was less about broad-based momentum and more about the concentration of capital, stage, and even geography. According to a recent article examining data from Q1 of this year, Latin American startups raised a total of $1.03 billion for the quarter, a jump of 12% YoY and a decrease of 6% from Q4 2025.

To see what is currently driving VC investment in Latin America, let’s take a deeper dive into the Crunchbase data from Q1 2026 and their analysis.

Late-Stage Dominance

In the first quarter, late-stage funding dominated VC investment in Latin American startups. Crunchbase data shows that $761 million of that total went to late-stage and growth deals, a jump of 158% from Q1 2025 and 203% from Q4 2025.

At the other end of the spectrum, round counts and total funding were down across angel, seed, and early-stage rounds. In fact, less than 9% ($92 million) of the Q1 total went to angel and seed stages. Early stages brought in 17% ($179 million), again a drop QoQ and YoY.

The discrepancy in funding between stages indicates that investors are prioritizing scale, proven business models, and clear paths to profitability instead of early-stage experimentation and greater risk. Capital is still available across the board, but it has become increasingly selective, and those without traction or near-term scalability are finding that funding can be harder to come by.

Geographical Concentration

Mexico was the clear winner in Q1, due in large part to investment in one startup that brought in almost one-third of total funding. That went to Mexico-based Kavak with a $300 million Series F round. This helped boost the total for Mexican startups in Q1 to $404 million and put them ahead of Brazilian startups who raised $240 million.

This is a notable and historically rare shift in regional leadership. Brazil has traditionally been the leader for venture funding in the region, so this is a bit of a pivot, and according to Crunchbase, only the second time since 2012 that Mexican startups outraised Brazilian startups in a quarter.

In looking at the data collectively, it shows that Latin America’s venture market is not contracting, but rather maturing.  Investors are targeting fewer, higher-quality companies, and they are re-engaging at the top end of the market. It also shows that regional dynamics are becoming more fluid. All of this indicates that the bar has been raised for founders in the region, but for the right companies at the right stage, the market is still very much open.

Three of us spent last week in Beverly Hills for the Milken Institute Global Conference. The official sessions ran May 3 through 6 across the Beverly Hilton and the Waldorf, under the banner “Leading in a New Era.” As always, the more useful conversations happened in the side rooms, the hotel suite meetings, and the Tuesday night party we sponsored. A few things we are bringing back to clients.

Jensen on Monday

Jensen Huang’s fireside with Becky Quick was the hottest ticket of the day. Room was over capacity, long line down the Hilton hallway, no chance of getting in late. Three points stuck with us.

First, his framing of AI as “the five-layer cake” rather than a model. AI is infrastructure, not software, and you should be diligencing exposure at every layer of the stack. Useful frame for anyone underwriting an AI-adjacent target.

Second, his pushback on AI job-loss narratives. He drew a real distinction between automating a task and eliminating a role, and said the doom rhetoric (much of which, he noted, comes from inside the AI industry) risks making Americans disengage from the technology entirely. Clients who want us to load up employment reps with AI-displacement language should think hard about how much weight that actually carries.

Third, on geopolitics. He was clear that China should not have access to Nvidia’s most advanced chips, and that US companies have no business blocking lawful government use of AI for national security. The second part read as a direct shot at Anthropic’s stated posture. Worth tracking if you are advising dual-use clients on procurement work.

Fink and Flatt on Tuesday

The most useful session of the week was Larry Fink’s panel with Brookfield’s Bruce Flatt, moderated by Mike Milken. Fink’s headline call: there is no AI bubble. There is the opposite of one. The US is short power, short compute, and short chips. He put the capex need at roughly $10 trillion this decade and made the obvious follow-on point that governments running structural deficits cannot fund this themselves, which puts the build on the private sector. He also previewed a BlackRock partnership with an unnamed hyperscaler on a 1-gigawatt data center, to be announced later in the week.

The forward call worth flagging: Fink expects a new asset class to emerge in compute futures. Forward contracts on raw computing power, priced the way markets currently price energy and agricultural commodities. There is no regulated venue for that yet. But if BlackRock is teeing it up on stage at Milken, it is closer than most people think. Quietly an interesting derivatives, tax, and ERISA project for somebody.

For our funds clients, the more practical Fink point was on retirement capital. Over half of BlackRock’s $14 trillion is retirement money. The DOL’s March 30 proposed rule on private credit access for defined contribution plans is the next real battleground. Fund managers raising vehicles built to tap that channel should be in front of ERISA counsel now, not later.

Flatt’s complement was useful in its own right. 50% of Brookfield’s $1.2 trillion private credit book sits in asset types that were not investable 15 years ago. He expects that share to reach 75% in another ten. The covenant packages we are reviewing right now reflect that shift. Tighter than the marketing decks would suggest.

The IPO window

Quick note for our growth-stage founders. The window is open. Most of you do not fit through it. The cohort actually getting out is small, profitable, and AI-adjacent. Most everyone else is being told to come back in 2027 with two more clean quarters. Dual-track, keep your M&A optionality alive, and do not burn cycles drafting an S-1 you cannot defend on a roadshow. Strategics are still paying better multiples than the public market for anything sub-$200M ARR, and that gap is wider than the headlines suggest.

The TIDE Mansion Party, with Grace and Baron

The highlight of our week was Tuesday evening at the TIDE Mansion Party, hosted by Grace Reyes and the team at The Investment Diversity Exchange together with Baron Davis and Business Inside the Game. Â鶹´«Ă˝ was proud to return as a sponsor this year, alongside featured sponsor Safra National Bank and our fellow sponsors Punch Financial, Qapita, and 645 Ventures. The co-host roster reads like an LP and emerging manager directory in its own right, from Franklin Templeton to Waterfall Asset Management to Deerpath Capital, which gives you a sense of who was actually in the room.

The party landed on Cinco de Mayo and Grace and Baron’s team leaned into the date. Sombreros, cactus, the works. Off the record, invite-only, a few hundred senior institutional investors and the GPs they actually take meetings with. Different conversation than the one happening on the Hilton stages, which is the point. Several of our most productive client introductions of the week happened on that patio. Thank you to Grace, Baron, and both teams for putting us in that room again. The follow-ups are already flying.

Bottom line

Dry powder is at record levels and the bid-ask on quality assets is finally tightening. Deals that pencil are getting papered. They are just taking longer than they did 18 months ago, and the docs are doing more work. If you were in town and we missed you, drop a line. We are all back through LA later this quarter.

Louis Lehot, Brian Wheeler, and Gus Resendiz are partners at Â鶹´«Ă˝ & Lardner.

San Francisco, CA: Figure AI is now valued at $39 billion. Humanoid robotics funding grew roughly 15x in three years, to $3.7 billion in 2025. Tracxn counts 121 humanoid companies globally, sitting on $6.89 billion of venture capital.

Six weeks ago, we wrote about the inaugural Hard Things — a Palo Alto living room with Veena Radhakrishna of Cartesian Kinetics, in the same week Jensen Huang disclosed a $1 trillion Nvidia order book at GTC. Veena gave us the operator’s framework: design for retrofit, choose the intelligence layer over the hardware layer, stay flexible long enough for the market to clarify.

Last week in San Francisco, Round Two gave us the other half. Veena gave us the “how”. Andra Keay of Silicon Valley Robotics gave us the “why now.”.  She told us why a field unfundable for fifteen years has become the place every generalist VC is anxious about showing up late to.

Before It Was Cool

Andrea has run Silicon Valley Robotics, the world’s largest cluster of robotics and AI companies, since 2011. She also founded Women in Robotics and writes “Robotics and Startups” on Substack. If a robotics company in this Valley matters, Andrea probably saw it first.

Andrea arrived in Silicon Valley at the end of 2010 looking for a 20-year technology bet. She picked robotics. Then she had to find robotics. At her first SVR gathering, there were thirty companies. She mapped the rest of the local ecosystem and got to sixty. There was no venture industry for robotics in 2011.  There were just a few hardcore investors drifting in from semis, a handful of founders trying to spin tech out of Willow Garage, and that was it.

For the next several years she crashed every venture event she could. She got laughed at. “Hardware is too hard” was the standing line. By 2015, she was tracking about $1 billion of global venture investment in robotics, and in in hindsight, probably double that. One of many mic drops of the evening: when Andrea went back and looked at the early investors who told her robotics was unfundable, the most common credential in their bios was a fine arts degree.

Today the bios read differently. Physics PhDs. Hardware operating experience. NVIDIA’s NVentures alone has roughly $2.1 billion of cumulative humanoid investment. Strategic capital from CATL, JD.com, Alibaba, and Amazon is outpacing traditional VCs. The category that used to be a punchline is now the belle of the ball.

What Actually Changed

We asked Andrea what the biggest unlock has been. Foundation models? Hardware costs? Capital? None of those. The biggest unlock has been time and wisdom. Founders learned a scientific approach, defining what they were trying to prove, what milestones would prove it, what would force a pivot. Investors learned to underwrite hardware risk. The ecosystem matured. The technology, in her telling, has always been moving. What was missing was the ability to build companies on top of it that could actually ship.

This is the historical proof point underneath Veena’s framework. The market always clarifies later than the models predict and faster than the incumbents are ready for. Andrea walked us through fifteen years of exactly that pattern:  hype cycles running a year ahead of evidence, disillusionment burning a year after, real industries quietly compounding underneath both. The founders who survived built organizations flexible enough to keep learning while the market took its time.

It also vindicates the larger Round One thesis: “if AI is eating software, hardware and robots are at the top of the food chain.” The phrase is Andrea’s. Software gets easier to churn out of every quarter. Robotics gets “stickier” once it’s in, not because the metal is hard to rip out, but because the integration into operations and the digitization of three-dimensional process data creates a customer with nowhere to churn to. Every deployment is different. You cannot drop in a replacement.

The Bubble Isn’t What You Think

So inflection or bubble? Andrea’s answer: there’s a bubble, but it isn’t in robotics. It’s in capital.

Unitree, the world’s most prolific humanoid manufacturer, shipped about 5,500 units last year, and that may quietly include quadrupeds. Maybe ten companies sit at that production scale. There are two to three million industrial robots in the world today, after sixty years of industry development. The forecasts imagining ten billion humanoids by 2040 are, in Andrea’s word, “no way.” The realistic ceiling looks more like 500,000 units by 2035, and the market still hasn’t figured out what most of them will actually do for a living.

This is why the Figure-BMW pilot mattered, even at 50% uptime. Not a verdict on the technology, just a useful piece of discovery. Every humanoid deployment for the next five years should be read as discovery, not validation. Same lesson Veena pushed in Round One: let customer environments teach you what your roadmap can’t predict. Andrea gave us fifteen years of receipts.

The line that came up repeatedly: pick high-value use cases. Robotics fails chasing low-margin automation. It wins chasing labor that’s scarce, expensive, regulated, or impossible to attract.  Think pharmaceuticals with chain-of-custody, blood-sample transport, cold-chain medication, industrial work nobody under thirty wants. The opportunity is high-value cargo where someone is “already” paying a fortune for a credentialed human to do the job badly.

And — Andrea channeling Tolstoy — staked the view that happy startups are all alike. Stop asking the founders who succeeded she says, and ask the ones who failed at exactly what you’re trying to do. She called it “institutional graveyard” wisdom. We are stealing the phrase.

The Hand Problem

One of the most arresting moments came from the audience. Tapan, Chief Scientist at Adverb, made a point most generalist humanoid investors do not yet appreciate. The current race is overwhelmingly focused on vision and language models trained on internet-scale data. But you cannot light a match with anesthetized fingers, no matter how many tokens you’ve seen. Without tactile feedback (think pressure, force, grip), robots cannot do the things humans do with their hands. And humans do most useful physical work with their hands.

Andrea agreed. The current AI advances still operate largely on abstractions of physical-world action. The missing layer is embodiment feedback. The most exciting near-term development she sees isn’t a humanoid company at all. It’s the wave of new sensor and tactile-grasp companies. “Hand as a service,” she joked. We aren’t sure she was joking.

Your First Household Robot Is Your House

Asked what non-obvious scenario the world is missing five to ten years out, Andrea said: your first household robot isn’t going to be Rosie. It’s more likely going to be your house.

The future, in her telling, isn’t a parade of humanoids. It’s intelligence quietly embedded in furniture, fixtures, vehicles, infrastructure, hospitals, fields. The average car is already a robot. Most of the robotics we live with in 2035 will be invisible. The flip side: judgment, presence, taste, relationships (the things most valuable about being human) become more economically valuable, not less. Robot chefs were supposed to obliterate restaurant work. People still pay more for food cooked by a person.

The Through-Line

Step back and the picture is consistent. Round One gave us the founder’s discipline: stay flexible, design for retrofit, choose intelligence over hardware. Round Two gave us 15 years of evidence that this is the only discipline that has ever worked in robotics — and the data showing the field is finally ready to reward the founders practicing it. Veena was the supply signal. Andra was the ecosystem signal. Physical AI is no longer a thesis. It is a capital allocation decision.

Practical Implications

If you are “building,” treat the two posts together as the operating manual. Pursue high-value use cases. Make pilot customers pay.  Customers need to have skin in the game, even if not full price. Plan financials both ways, since only about 20% of any segment actually wants RaaS. Study founders who failed at your exact use case more than the ones who succeeded.

If you are “investing,” watch headcount growth, not press cycles. Watch order books, not pilot announcements. Underwrite founders who articulate what they expect to “learn” from each step. Most robotics failures are financial and operational, not technical.

If you are an “enterprise buyer,” treat early humanoid pilots as data-gathering. Focus on problems no one is solving today because the labor doesn’t exist or doesn’t want the job.

For help on the legal, capital, and strategic dimensions, Â鶹´«Ă˝ & Lardner and Mavka Capital are here. Reach out.

One More Thing

Hard Things Round Two did what we wanted. A small room. A speaker who has paid the price for her opinions. The kind of conversation that doesn’t happen on a panel. We are building, evening by evening, a serialized field guide for the next decade of Physical AI from operators who lived through the wilderness years and stayed. Round Three is coming.

Thank you to the Mavka Capital and Â鶹´«Ă˝ teams who make this series happen. Thank you to Tapan for the hand-problem moment we are still chewing on. And thank you to Andrea for reminding the room that the field everyone wants to fund right now is the same field almost nobody believed in 15 years ago. The fundamentals haven’t changed. The capital has finally caught up.

See you at the next one.

Further Reading

• “A $1 Trillion Order Book and a Palo Alto Living Room: What GTC 2026 and Hard Things Taught Us This Week About Physical AI (https://louislehotattorney.substack.com/p/a-1-trillion-order-book-and-a-palo)” — Hard Things Round One, with Veena Radhakrishna of Cartesian Kinetics
• “Robotics and Startups” by Andrea Keay on Substack
• “The Moat Is Dead. Long Live the Runner” by Vitaly Golomb

About the authors

Louis Lehot is a veteran Silicon Valley emerging growth technology company lawyer at Â鶹´«Ă˝.  Vitaly Golomb is the Managing Partner of Mavka Capital and Mavka Ventures, an early-stage embodied AI fund, and author of “Accelerated Startup.” Together, they co-host the Hard Things invite-only event series. For legal, fundraising, M&A, or strategic advisory on physical AI and robotics, reach out to Louis or Vitaly directly.

Key Takeaways:

• Investment in AI infrastructure presents a compelling opportunity for private equity (“PE”) funds and is increasingly becoming a target of investment.
• Brookfield estimates that $7T will be spent on AI-related infrastructure over the next 10 years, extending beyond just data centers.
• Firms building out portfolios of interconnected companies powering AI must consider structural, regulatory, and financial issues to ensure success.

The infrastructure powering artificial intelligence (AI) is equally as important as the technology itself. What began as a rush to invest in the latest AI startups has now evolved into something more significant: investing in the development of the fully integrated infrastructure systems that enable AI to operate at scale. Leading PE firms have already been pouring , transforming them from a niche investment to a staple of their portfolios.

To put it in perspective, a report from estimates that total spending on AI-related infrastructure could exceed $7 trillion within the next 10 years. This investment is broken down into the following categories:

• $2 trillion for developing new data center capacity
• $0.5 trillion for baseload power and electricity transmission infrastructure
• $4 trillion for GPU partnerships and chip design/manufacturing
• $0.5 trillion for dedicated fiber connectivity, cooling solutions, and semiconductor and robotics manufacturing

AI does not rely solely on data centers, creating significant opportunities for investment beyond just the land acquisition and facility construction. AI requires massive reliable power inputs, high-speed data transmission, and localized processing capabilities. This is prompting PE firms to expand into adjacent sectors that were previously outside their typical investment scope.

For example, investing in energy solutions is a critical component of AI infrastructure strategies. Similarly, there is a renewed interest in fiber networks and connectivity assets, as efficient data movement is essential for AI applications, especially those requiring real-time processing. Edge computing has also increased the need for processing power closer to the point of use, driving investment into smaller, distributed facilities and edge nodes that complement centralized data centers.

Because there are so many components making up the infrastructure behind AI, it opens the opportunity for a new platform strategy for PE firms. It allows firms to build a vertically integrated portfolio of companies that span power, connectivity, and compute, with multiple acquisitions across sectors and long-term development pipelines, all with the goal of delivering a unified service offering.

This integrated platform approach raises governance and structural considerations for sponsors as they consider how to allocate capital across various strategies, manage conflicts between funds, and structure investments. Co-investment vehicles, joint ventures, and continuation funds are viable options, but each presents its own challenges.  An integrated platform also comes with its own level of risk as one issue in one layer of the platform can have a ripple effect throughout the platform. As a result, more sophisticated structuring of intercompany agreements, supply chain arrangements, and contingency planning is required.

There is also a great deal of regulatory consideration that come with these industries. Energy, telecommunications, and data-related assets are all subject to a complex web of regulations and compliance requirements and combining them in a single platform multiplies the regulatory burden. The capital intensity of integrated AI infrastructure is also significant, and the cash flow profiles of these different asset classes can vary widely. To blend all of these into one cohesive financing structure requires innovative thinking.

Demand for AI infrastructure is not only growing; it is becoming increasingly interconnected. To capture the immense value, sponsors must be able to assemble and operate integrated platforms that offer end-to-end solutions. This approach supports premium valuations and creates multiple exit pathways. It’s about moving beyond the data center to build comprehensive systems that will power the future.