Hello all,
There is a provocative piece making the rounds lately from Mavka Ventures, a VC firm, titled “The Moat is Dead. Long Live the Runner.” The thesis is bracingly simple: in the age of AI, competitive moats — the protective advantages that Warren Buffett spent a career identifying and investing in — are obsolete. The future, we are told, belongs not to castle builders but to speedboat captains. Velocity of execution is the new everything. The thickest walls no longer matter; only the fastest legs do.
It’s a compelling narrative. It’s also only about half right.
As someone who has spent over 30 years in the intellectual property market — buying, selling, licensing, and valuing patents — I have watched more than a few “the old rules are dead” manifestos come and go. Some had merit. Most ended up being the business equivalent of declaring gravity optional because you just invented a really fast elevator. So let’s take a fair, balanced look at this one. Because while the speed argument has real force, the case for the moat remains far more durable than its critics suggest. And in the IP world, the answer turns out to be refreshingly simple: you need both.
As usual, while I focus on the macro picture here, we track everything happening in this space. For a more frequent news fix, you can follow me on LinkedIn, where I post almost daily on noteworthy developments. Links to recent posts are available here.
Happy reading!
Louis
__________________
Tangible IP News
After nearly a decade with Tangible IP, Erika Warner will be leaving at the end of the month. Erika has been an integral part of our team, contributing to dozens of patent transactions and building strong relationships with clients and industry partners along the way. We are deeply grateful for her dedication and professionalism over the years and wish her every success in her next chapter.
“My time at Tangible has been deeply meaningful—professionally and personally. I’ve had the privilege of working alongside talented colleagues, innovative clients, and trusted partners across the IP ecosystem. Together, we’ve navigated complex transactions, built lasting relationships, and contributed to a dynamic and evolving market. I’m especially grateful for the opportunities to grow, to lead, and to collaborate with so many thoughtful and driven people along the way. While it’s never easy to leave a place that has been such a significant part of my journey, I wish Tangible IP nothing but continued success, and I carry forward a deep sense of gratitude for the role it has played in my career,” remarked Erika Warner.
Business continues as usual at Tangible IP — our team remains fully engaged and committed to delivering the same high-quality IP brokerage and advisory services our clients have come to expect over the last 15 years. In this regard we are delighted to report we recently brokered the sale of a Marketplace with Integrated Trust Networks patent portfolio. Details on recently completed transactions can be found here.
Louis Carbonneau was also recently a guest on Gene Quinn’s Unleashed podcast discussing, The Patent Monetization Problem: How Innovation Without Return Distorts the Market. You can revisit the conversation here.
__________________
Featured Portfolio
Tangible IP is pleased to represent Lumanary, Inc., in the divestiture of their patent portfolio in the Automated Video Content Management and Distribution domain. This portfolio forms a cohesive suite of innovations centered on core functionality behind modern video platforms including automatic insertion of branded bumpers, intros, watermarks, and metadata-driven publishing across third-party platforms. The technology addresses critical workflows adopted by major social media platforms – including video hosting, cross-platform content syndication, and automated repurposing technologies embedded in mainstream social media and streaming operations.
Please reach out to info@tangibleip.biz to receive the marketing materials and additional information.
__________________
Which Is Better: A Speed Boat or a Moat?
Let’s give credit where it’s due. The Mavka piece correctly identifies a tectonic shift in the tech landscape: the cost and time required to build software products has collapsed. AI wrappers built on commoditized foundation models are shutting down at accelerating rates because anyone with an API key and a long weekend can replicate them. In 2025 alone, AI startups captured nearly two-thirds of all U.S. venture capital, and that flood of money didn’t create defensibility — it destroyed it. When your six-month head start can be cloned in a two-sprint sprint (their phrase, and a good one), you’d better be running fast.
The piece also nails a structural shift in competitive dynamics: big tech no longer has the decency to be slow. Google, Microsoft, Amazon, and Apple now have venture-scale speed with Fortune 10 resources. The old startup playbook — find a niche, grow before the giants notice — assumed a window of inattention that has essentially closed. And as if that weren’t enough, an ironic twist: the FTC’s aggressive posture on acquisitions has actually made things worse for startups. Rather than acquiring promising companies (which at least provided exits), incumbents now simply build competing products internally. Startups get neither a buyer nor breathing room. (One might call this the regulatory equivalent of locking the fire exit and then complaining about the smoke.)
So yes, speed matters. Enormously. Execution velocity has always been important, but in an era when the cost of building has cratered, moving fast may be the only thing standing between you and irrelevance. Point taken.
But here’s where I get off the speedboat!
The Case for the Moat: Where the Obituary Is Premature
The fundamental problem with the “moats are dead” argument is that it confuses a particular type of moat — one built on technical complexity in software — with the concept of moats itself. Saying software moats have eroded does not mean all moats have evaporated. That’s a bit like saying “locks don’t work anymore” because someone invented a better lockpick. The correct response is to get a better lock — not to leave your door wide open and hope you can outrun the burglars.
Let me be blunt: if speed is your only moat, you don’t actually have a moat. You have a head start. And head starts, by definition, are temporary. As CXL put it rather elegantly, the difference is like a mid-season trade (a short-term roster boost) versus a strong farm system (which delivers year after year). Mid-season trades are competitive advantages. Farm systems are moats. Speed gets you to market first. A moat is what keeps you there.
This is not an abstract point. Morningstar’s research has consistently shown that companies with wide economic moats — those with defensible advantages expected to last 20 years or more — dramatically outperform those without. When Buffett said he looks for “economic castles protected by unbreachable moats,” he wasn’t being quaint. He was describing a framework that has produced the most successful investment track record in modern history. And when Elon Musk dismissed moats as “lame” in their now-famous 2017 exchange, Buffett’s reply was characteristically dry: “Elon may turn things upside down in some areas, but I don’t think he’d want to take us on in candy.” (For the record, See’s Candies — a Berkshire Hathaway holding with one of the widest brand moats in consumer goods — has been printing money for decades, and it has never shipped a single software update.)
Even Charlie Munger acknowledged that “old moats are getting filled in and new moats are harder to predict.” But “harder to predict” is not the same as “non-existent.” It simply means the nature of moats is evolving — not disappearing.
Intellectual Property: The Moat That Refuses to Die
Now, here’s where things get particularly interesting for our audience. If you’re reading this newsletter, you likely work in or around the IP ecosystem — patents, licensing, technology transactions. And in this world, the notion that moats are dead is not just wrong; it’s almost comically wrong.
Consider the pharmaceutical industry. Drug patents remain perhaps the most powerful moats in existence. A single well-crafted patent can protect billions in annual revenue for up to 20 years. The pharma sector is currently staring down a patent cliff of between $230 and $400 billion in revenue at risk by 2032 as blockbuster drug patents expire — on products like Merck’s Keytruda, Bristol Myers Squibb’s Eliquis, and Novo Nordisk’s Ozempic. The fact that these companies are frantically spending tens of billions on M&A to replenish their pipelines tells you everything you need to know about the value of a patent moat. If moats didn’t matter, nobody would be paying $10 billion in a bidding war for a clinical-stage biotech (as Pfizer did for Metsera in late 2025). That’s not a company buying speed. That’s a company buying a moat.
And here’s the delicious irony: pharma is an industry that absolutely requires both speed and a moat. Drug companies race to get treatments through clinical trials and regulatory approval (speed), but without patent protection (the moat), the entire economic model collapses. Why spend $2.2 billion developing a new drug if competitors can copy it the moment it hits the market? Speed gets you to the FDA first. The patent is what keeps generics at bay for the next two decades. Take one away and the whole system falls apart — like trying to row a boat with only one oar. You’ll go in circles, which, come to think of it, is a perfect metaphor for speed without a moat.
The same logic applies across the technology landscape. In the AI patent wars, just four companies — Google, Microsoft, IBM, and OpenAI — hold over 20,000 AI-related patents. Microsoft alone has more than 500 patents covering neural network optimization. These aren’t decorative moats. They’re functional ones, providing legal leverage that prevents competitors from freely replicating core technologies. Meanwhile, AI patent filings have surged by more than 30% annually, and startups account for 35% of new U.S. AI patent filings — which tells you that even in the fastest-moving technology space on earth, smart companies are still building moats, not just speedboats.
The Analogy Problem: Speed Without a Moat is a Treadmill
Let me offer a few analogies that illustrate why the “speed is everything” thesis breaks down when you push it to its logical conclusion.
Consider professional sports. The fastest player on the field is not always the most valuable. Usain Bolt was the fastest human alive, but you wouldn’t draft him to play quarterback. In the NFL, speed matters enormously — but so does scheme, playbook complexity, team cohesion, and institutional knowledge. Speed without strategy is just chaos with better cardio. The teams that win championships are the ones that combine talent acquisition velocity (drafting, trading, free agency) with structural advantages (coaching systems, culture, front-office intelligence). That’s the moat-plus-speed model.
Or consider restaurants. A new restaurant that opens fast and grabs attention with a brilliant first menu will get press coverage and a packed opening week. But without the moat — a distinctive culinary identity, a loyal customer base, proprietary recipes, a location advantage, a brand — it will be displaced by the next shiny object within a year. Speed gets you the Yelp review. The moat gets you the 20-year reservation list. (Ask any restaurateur whether they’d rather have a viral TikTok moment or a Michelin star that keeps renewing. The answer won’t surprise you.)
Or, for a more tech-adjacent example: consider Amazon. Bezos famously prioritized speed of execution and customer obsession. But he also built one of the most formidable moat ecosystems in business history — a logistics network that would cost over $100 billion to replicate, a cloud platform (AWS) with massive switching costs, a Prime ecosystem that locks in consumers with compounding benefits. Amazon was fast. But Amazon also built walls. The speed without the walls would have produced WeWork — a $47 billion valuation that sprinted its way straight into bankruptcy.
The Real Framework: Speed + Moat = Durable Advantage
Here’s where I think the Mavka piece actually gets tantalizingly close to the right answer before veering away from it. They acknowledge that “speed alone isn’t enough” and that the real alpha lies at the intersection of speed and “non-obvious insight.” That’s a repackaging of the moat concept — they just don’t want to call it that. “Domain expertise that others lack”? That’s an intangible asset moat. “Deep, earned insights that come from years of operating”? That’s proprietary knowledge — also a moat. They’re arguing against moats while simultaneously describing them, which is a bit like a fish writing an essay about the irrelevance of water.
A more honest framework would acknowledge what McKinsey’s 2025 AI research found: 79% of organizations report that competitors are making similar AI investments, yet only 23% believe they’re building sustainable advantages. The gap between investment and differentiation has never been wider. If speed alone created durable advantage, that gap wouldn’t exist. The companies pulling ahead are the ones that combine execution velocity with assets competitors cannot replicate — proprietary data, network effects, deep workflow integration, and yes, intellectual property.
The winning formula in 2026 is not moat or speed. It’s moat and speed. Think of it this way: speed determines how quickly you can seize an opportunity. The moat determines how long you can hold it. A company that moves fast but builds no structural defenses is essentially sprinting on a treadmill — lots of energy, lots of sweat, zero forward progress. A company that builds a moat but moves slowly gets disrupted before the moat fills with water. You need both. The castle and the cavalry. The wall and the horse.
What This Means for the IP Market
For those of us in the patent ecosystem, this debate carries enormous practical significance. If the “moats are dead” crowd is right, then the entire value proposition of intellectual property — creating legal barriers that protect innovation and reward R&D investment — is in jeopardy. But the data tells a very different story.
AI patent filings are exploding globally, with over 14,000 generative AI patent families filed by 2023 alone, up from just 733 in 2014. Companies are pouring resources into building IP-based moats precisely because they understand that speed without protection is a losing game in the long run. In our own brokered patent market, we continue to see robust demand for quality portfolios — particularly in AI, cloud computing, 5G/6G, autonomous vehicles, and cybersecurity — because buyers understand that patents remain one of the few competitive advantages that a well-funded competitor cannot simply clone by hiring more engineers.
This is the fundamental insight that the “speed-only” crowd misses: you can replicate a product, but you cannot replicate a patent. You can match a feature, but you cannot infringe a valid claim without consequences (or at least, you shouldn’t). You can outrun a competitor to market, but you cannot outrun an injunction. In a world where everything else can be commoditized, IP is one of the few remaining sources of genuine, enforceable differentiation. That’s not a relic of the old economy. That’s a moat.
So, what happens when you stop running?”
Because that’s the question the speed-only thesis never answers. Eventually, every sprinter slows down. Markets mature. Competitors catch up. Capital markets turn. And when that happens, you’d better have something more durable than velocity — something that holds your position even when the pace of execution levels off.
History is littered with companies that were fast but moatless. Groupon was lightning-fast to market. Vine moved faster than TikTok. Myspace was the king of social networking before Facebook even existed. Speed did not save them. What they lacked was a structural advantage that could survive the arrival of bigger, better-resourced competitors.
The real lesson of the AI era is not that moats are dead. It’s that moats must be earned faster, defended more actively, and built more creatively than before. The nature of the moat is changing — from static walls to dynamic ecosystems of IP, data, brand, and network effects. But the principle endures: if you want to build something that lasts, you need a defense that outlasts you. So by all means, build a faster boat. But don’t forget that your boat won’t go anywhere if the moat is dry.
__________________
Squires Moves to Close the Reexam Backdoor – But Is It Enough?
On April 1, Director Squires issued an Official Gazette notice establishing a new “Pre-Order Procedure” for ex parte reexaminations, effective April 5, 2026. For the first time, patent owners may file a 30-page paper (no petition or fee) within 30 days of service, arguing that the cited references do not raise a substantial new question of patentability. The requester generally cannot respond unless the patent owner’s paper contains misrepresentations — and even then, only via a 10-page petition with fee.
The context is a reexam filing explosion driven directly by the Squires discretionary denial regime: Deputy Director Stewart confirmed at IPWatchdog LIVE that requests are on track to exceed 1,000 in FY 2026, double the FY 2025 total.
The playbook is simple — if your IPR petition is denied (or heading south), repackage the same art into an EPR and refile at the CRU, which has been granting these at a 95%+ clip. The Federal Circuit’s December 2025 decision in In re Gesture Technology Partners made things worse by confirming that IPR estoppel does not extend to ongoing ex parte reexams. The timing gamesmanship was laid bare in Western Digital v. Godo Kaisha IP Bridge, where the petitioner filed an ex parte reexam within 48 hours of an unfavorable companion FWD — using art that had been available all along — specifically to beat estoppel attachment on the remaining IPR.
Will the new procedure help? At the margins, yes — it should bring the grant rate down and knock out the weakest requests early. But as Gene Quinn noted in a same-day IPWatchdog piece, the real mischief is not SNQ-related: it is the timing-based estoppel avoidance that the pre-order procedure does not reach, since patent owners are specifically barred from raising discretionary considerations. Until Congress or the Federal Circuit moves estoppel attachment from the FWD to institution, the CRU will remain a backdoor for determined challengers — and patent owners’ PTAB victories will stay provisional at best.
__________________
The Government Wants eBay Dead – And It’s Not Being Subtle About It
Twenty years after eBay v. MercExchange effectively turned the patent right to exclude into a judicially administered compulsory license — particularly for NPEs — the federal government has apparently decided it’s had enough. In the span of just twelve months, the USPTO and DOJ Antitrust Division have filed three joint statements of interest arguing for broader injunctive relief, after filing virtually none in the prior two decades. If that doesn’t qualify as a policy reversal, I have a bridge in Brooklyn to sell you.
The most recent — and most consequential — was filed on February 27, 2026 in Collision Communications v. Samsung, where Judge Gilstrap is deciding whether to grant a permanent injunction to Collision, an NPE that won a $445.5 million willful infringement verdict over signal-processing patents used in Samsung’s 4G, 5G, and Wi-Fi products. Samsung’s defense? The usual post-eBay circular logic: as an NPE, Collision expected to license for royalties, so money should be adequate. (Translation: “We refused to take a license, infringed willfully for years, and now we’d like to argue that royalties are all the patent owner ever wanted.”)
The government’s joint statement went straight at this reasoning, arguing that NPEs can demonstrate irreparable harm and inadequacy of damages where the patent is difficult to value and infringement damages hard to calculate. The USPTO press release didn’t mince words: “Non-practicing patentees should not be categorically denied the opportunity for injunctive relief.” This follows two earlier joint statements — one in Radian Memory Systems v. Samsung in district court, and another at the ITC — each pushing the same theme. Three bites at the apple in twelve months, across two different forums. That’s not a coincidence; that’s a coordinated campaign.
And it fits squarely within the broader posture we’ve been tracking in this column. The same USPTO that is now arguing for NPE injunctions has simultaneously dramatically increased discretionary denials of PTAB petitions and driven institution rates to historic lows. Taken together, the message to the judiciary could not be clearer: the executive branch wants patent rights intact and enforcement broad — even for entities whose business model is licensing rather than manufacturing. For those keeping score at home, this is the most pro-patent posture from a U.S. administration since… well, possibly ever.
Now, to be fair, the government’s legal arguments still have some gaps. As one pundit noted in a sharp analysis, the uncertain-patent-value theory came without a single citation to Federal Circuit patent authority, and the cases cited for the irreparable-harm proposition all involved parties that competed in the marketplace — not pure licensing entities. So the doctrinal scaffolding isn’t fully built yet. But let’s not confuse legal craftsmanship with political signal. When the government shows up uninvited in three separate cases to say the same thing, courts notice — even if the brief could use a few more footnotes.
Meanwhile, the RESTORE Patent Rights Act (which we’ve covered before) continues to gather bipartisan support in both chambers, offering the legislative backstop: a rebuttable presumption of injunctive relief upon a finding of infringement. Between the executive-branch statements and the legislative momentum, the post-eBay order is being squeezed from both ends.
What It Means for Patent Values
This is where it gets interesting for anyone buying, selling, or financing patents. The economics are simple: a patent’s value in negotiation is a function of the consequences the other side faces if it doesn’t take a license. Under the current regime, those consequences amount to paying a court-determined royalty — roughly what a pre-infringement license would have cost. Restore a credible injunction threat, and you restore the multiplier effect that used to drive patent transactions. (Remember the pre-eBay BlackBerry case? A $53.7 million verdict became a $612.5 million settlement — eleven times the award — because the injunction was real.)
If injunctions become credible again, expect licensing rates to rise across the board, settlement multipliers to increase, litigation funding to become more attractive (higher expected returns per case), and — most significantly — the pre-litigation licensing market to revive. That last point is the big one. Most patent transactions happen in private, and their terms are shaped by the shadow of what might happen in court. Restore the shadow of an injunction, and you reprice every negotiation — not just the ones that go to trial.
We’re not there yet — Big Tech will fight this with everything it has, and Judge Gilstrap’s ruling in Collision could go either way. But the direction of travel is unmistakable, and for patent owners, this is the most consequential development in enforcement since eBay itself. Time to pay attention.