A recent analysis on DrugPatentWatch works through the loss-of-exclusivity histories of Revlimid, Humira, Eliquis and Keytruda and finds no cliff anywhere in them. Exclusivity ends in steps. A patent term extension here. A secondary or formulation patent there. A settlement date that matches no patent’s printed expiry. A regulatory exclusivity running on its own clock. Humira’s compound patent expired in December 2016. The first US biosimilar launched in January 2023. A forecast keyed to the compound patent was out by more than six years.

Why the exclusivity timeline beats the patent expiry date

The steps are the interesting part, because they are built rather than granted. Revlimid’s generics entered under volume caps fixed by settlement, not by any patent. Eliquis produced two different entry years for the same drug, years apart, depending only on whether a challenger settled or kept litigating. Merck is building a subcutaneous formulation ahead of Keytruda’s core expiry, and that defensive move has already drawn a fresh infringement fight of its own. Every step traces back to a decision someone took years earlier about what to file, what to extend, what to settle and what to reformulate.

Three things follow, and none of them are specific to pharmaceuticals:

– Model the timeline, not the date. What matters commercially is when competition arrives at full strength, and that number is nowhere on the front page of the patent. The Australian extension record makes the same point from the other side, as I set out in The Pharmaceutical Patent Cliff Has a Timetable — And It’s Public: many molecules are defended by three or more separately extended patents, and those dates are a published calendar of when each competitor becomes vulnerable.

– Treat the settlement date as the operative date, and structure it with that in mind. It usually controls, it can fall before the last patent expires, and it carries a long tail of risk – the lesson of The Reverse Payment You Didn’t Know You Made, where the economics of a deal signed in 2014 were still being unwound more than a decade later.

– Build the second layer while the first still has years to run. Formulation, method-of-use, device and delivery rights are what turn one date into several. The discipline is claiming every commercially viable variant before a competitor carves around you, as When Your Strongest Asset Is What You Leave Off the Label shows.

A step you negotiated is also a step someone can take from you. A listed patent is not a tested one, a settlement can be reopened, and a reformulation can invite a new opponent with a portfolio of its own.

So the useful question for your next board paper is not when the patent expires. It is what your exclusivity timeline actually looks like, which steps you built deliberately, and which ones you are relying on without ever having decided to.

Read the article: The Patent Cliff Is a Myth. What Actually Happens Is a Patent Staircase, and Most Models Miss the Steps, DrugPatentWatch.

Start with Twitter. A Virginia startup called Operation Bluebird has launched Twitter.now, reviving the old blue bird branding under the argument that X Corp abandoned the Twitter and Tweet trademarks when it rebranded in 2023. X disagrees, and is suing in Delaware federal court. What makes this more than a curiosity is a Delaware judge’s comment from the bench earlier this year that X “appears to have abandoned” its rights in Tweet and the bird logo, and possibly the word Twitter itself. No written order has issued, and the case is unresolved. But the lesson doesn’t wait for a verdict: trademark rights are use-based, not sentiment-based. If your business retires a name, a logo, or an old product line, someone else can start building a case that you gave it up – and “we didn’t mean to” is not a defence courts take seriously. Decide deliberately whether you’re keeping a mark alive, and keep evidence of continued use if you are.

Second, Huawei and HP have signed a multiyear global patent cross-licensing deal covering Wi-Fi technology, including the new Wi-Fi 7 standard. This resolves a dispute that started in 2025 when Huawei sued HP over Wi-Fi 6 patents at the Unified Patent Court. Nobody walked away from anything here – HP is paying for continued access, and both sides now hold reciprocal rights to each other’s portfolios. It’s a clean example of patent litigation doing exactly the job it’s designed for: turning an infringement standoff into a priced, ongoing commercial relationship. If your products touch standardised technology, you are already inside someone else’s licensing model, whether you’ve mapped it or not. The interesting move is asking, before a dispute forces the question, whose patents you’re already relying on and what a licence would cost if they came knocking first.

Third, the EUIPO’s Board of Appeal has confirmed that a well-known brand can defeat a copycat purely on reputation, without the earlier mark even needing to cover the same goods. Rovio successfully cancelled a gambling-industry trade mark that echoed Angry Birds’ colour scheme, character design and “crashing birds” concept, on the basis that the applicant’s dishonest intent could be inferred from the strength of Rovio’s reputation alone. This is a materially easier enforcement path than most brand owners realise exists: you don’t need a matching registration if your reputation does the work instead. It also means the return on building genuine brand recognition compounds in ways beyond marketing – it becomes a legal asset in its own right.

The pattern across all three: IP rights are shaped by what you actually do, documented and provable, not by what you intended. X Corp’s intentions about Twitter don’t matter if its conduct reads as abandonment. Huawei’s and HP’s dispute didn’t need to end in court once both sides priced the alternative. Rovio’s decades of visible brand-building is now doing enforcement work that no registration alone could achieve.

None of this is really about trademarks or patents specifically – it’s about whether your IP strategy is a record of decisions you actually made, or a set of assumptions nobody has tested. The three questions worth asking this week:

(1) What have we stopped using that we haven’t formally decided to abandon?

(2) Whose technology are we relying on without a licence? and

(3) Is our own reputation strong enough to do enforcement work if a registration alone won’t cut it?

Related reading: Why IP Strategies Fail: The Gap Between the Decision and the Portfolio, The $60,000 Business Name: Why an ASIC Search Is Not a Trade Mark Search and Who Gets Paid for the Platform? Lessons From a Big Week in AI and Intellectual Property.

Sources for this post

1. Twitter.now / X Corp v Operation Bluebird – TechCrunch, 27 August 2026; primary docket at CourtListener, X Corp. v. Operation Bluebird, Inc., 1:25-cv-01510.
2. Huawei/HP Wi-Fi patent cross-licence – Huawei press release, 26 August 2026; South China Morning Post commentary.
3. EUIPO Angry Birds bad faith decision – EUIPO Board of Appeal decision R1791/2025-4; The IPKat, 27 August 2026.

Writing in Harvard Business Review on 27 July 2026, executive coach Karen Walker gives this condition a name: ambient urgency. It is the chronic, low-grade state in which everything feels critical, nothing can wait, and decision quality quietly degrades. Her prescription is not another prioritisation framework. It is three tests to run before committing your attention: the replacement test (am I making this decision because only I can make it, or because I am the default?), the compounding test (will this create value that grows over time, or does it just relieve today’s pressure?) and the fear test (am I deferring this call because it is wrong, or because I don’t want to face its consequences?).

What the three tests look like in IP portfolio management

Run those tests against your portfolio and the results are usually uncomfortable. Renewal decisions made by habit fail the replacement test – as I argued in Why IP Strategies Fail: The Gap Between the Decision and the Portfolio, most companies’ IP spending correlates more than 90% with last year’s, which means nobody is really deciding at all. Invention harvesting, inventor relationships and trade secret capture pass the compounding test, and they are exactly the work ambient urgency squeezes out first – the slow erosion I described in The Quiet Decay: Why IP Value Slips When No One Is Watching. And portfolio pruning fails the fear test almost everywhere. Abandoning a granted patent or dropping a jurisdiction reads as an admission that an earlier decision no longer holds, so a company keeps paying renewals in fourteen countries for a product line it exited three years ago. Three questions are worth putting to your next patent portfolio review or IP audit:

– Which IP decisions genuinely need you – the patent-versus-trade-secret calls, the enforcement calls, the direction-setting – and which are you making by default?

– What in this year’s IP budget compounds, and what merely keeps the machine turning?

– What are we still protecting only because ending it would feel like failure?

The test of clear allocation is the same one that applies to any capital decision – the discipline I explored in The Bet You Make Before You Make Any Bet. An IP strategy is a record of trade-offs, made consciously. If your portfolio only ever grows, urgency is making your decisions for you.

Read the article: 3 Questions to Pressure-Test Your Priorities, Karen Walker, Harvard Business Review, 27 July 2026.

That reframing changes what is worth protecting and how. Note that the mechanisms Thompson identifies as durable – inference and memory efficiency, batching and caching, token efficiency, the harness the customer actually works in, and the data flywheel that comes from running inference at scale – are precisely the things that patents, trade secrets and well-drafted contracts can hold, in a way that published weights never could. Note too that the fight over distillation is a licensing fight, not a patent fight. It is being run through terms of service, which means it depends on contract drafting, on jurisdiction and on whether you can realistically enforce against the counterparty. This is the decision I have described elsewhere as choosing, in advance, which right to assert, where, and to what end.

Four questions follow, and none of them are about AI specifically:

– Ask what remains proprietary if a competitor obtains your core capability tomorrow. If the answer is nothing, the moat was rented.

– Check whether your operational efficiencies – the serving, caching and pipeline work nobody outside the team can see – have been captured as filings or as controlled trade secrets, or whether they are simply undocumented practice.

– Read the licence, not the label. Open-weight is not open source. Field-of-use limits, acceptable-use terms and downstream obligations travel with the weights into your product, and they are a freedom-to-operate question for your business, not a procurement footnote.

– If a supply contract is doing your protective work, confirm it is enforceable where the counterparty actually sits. A term of service that cannot be enforced is a preference, not a right.

The wider pattern is one that shows up well outside AI. Whenever a capability becomes broadly available, value moves to cost position, to proprietary data and to whoever owns the customer relationship. Those are IP decisions, taken early, or they are not taken at all.

Read the article: Who’s Afraid of Chinese Models?, Ben Thompson, Stratechery, 20 July 2026.

The IP signal sits in the remedy. Sector R&D runs at less than half a per cent of revenue, against 1.5 to 2.0 per cent in China, so the report advises companies to “forge global technology alliances” in order to “access frontier intellectual property”, and to pursue “targeted cross-border M&A and strategic investments in niche technologies”. That is a technology licensing market forming in public. If you hold patents in silicon carbide devices, cell chemistry, thin-film photovoltaics or high-voltage cable processing, a new class of licensees and acquirers is being told to come and find you. If you are the one scaling, you are being told to buy a position you do not own.

The durable lesson is in the title. Capacity is not a competitive position. Five times the plant, built on technology someone else owns, buys volume at someone else’s margin, and it fixes a royalty negotiated from the weakest possible seat, because leverage moves to the licensor the moment your capital is committed. Three disciplines follow, and none of them are specific to India:

– Price the in-licence before you commit the plant, not after.

– Match your patent filing strategy to the ten-year demand map rather than the historical sales map. A patent only works where it is in force, and import substitution is a plan to manufacture your technology somewhere you may never have filed.

– Treat the acquisition of niche technology as an IP due diligence question first and a manufacturing question second. What you are buying is a set of rights, encumbrances and freedom-to-operate assumptions, not a production line.

Export ambition raises the same question again. Certification, standards and testing are the gate to European and North American tenders, and that gate is where standard-essential patents and licensing obligations sit waiting. So the useful question for your next board paper is not how much capacity you are adding. It is whether the build, licence or acquire decision was taken before the capex was approved, or after.

Read the article: Wired for growth: India’s electrical-equipment opportunity, McKinsey and IEEMA, 21 May 2026.

The strategic work, then, is to treat the intelligence layer as an IP portfolio from day one. Three questions to consider now. First, ownership: when know-how is codified into agent skills — often with a vendor’s platform and people involved — do your contracts actually vest those skills, and the improvements from learning loops, in you? That engagement-and-ownership problem is the subject of AI Transformation Is Not a Strategy Problem — It’s an Ownership Problem. Second, secrecy: trade secret protection survives only with reasonable steps — access controls, provenance logging, and limits on what agents (and their providers) can see and disclose — which is precisely where autonomous systems create new leak paths, as explored in Your AI Agent Won’t Keep a Secret. Third, stewardship: compounding assets decay without active management, the pattern examined in The Quiet Decay: Why IP Value Slips When No One Is Watching. McKinsey warns of a “cognitive tax” — value flowing to the AI providers everyone depends on. The companies that avoid paying it twice will be those whose data, skills and know-how are identified, owned and protected as rigorously as any patent family. In the symbiotic enterprise, IP strategy isn’t just a legal workstream within the AI programme. It is the part that decides who keeps the value.

The strategic implication is that brand equity now has to be engineered to stand alone, not assumed as a by-product of distribution — the same lesson the Bodum design case taught in When the Monopoly Ends, Your Shape Has to Stand on Its Own: distinctiveness must be deliberately built and evidenced before you need it. Three questions worth putting on the agenda this quarter. First, if an AI intermediary described your product tomorrow, is the information ecosystem it draws on accurate, consistent and shaped by content you control — and are your registrations broad enough to act when it isn’t? Second, does your portfolio cover the formats growth is moving toward — experiences, licensing, characters, collaborations — or only the products you sell today? Third, as resale and customisation become mainstream, do you have a position on what third parties may do with your branded goods, an Nike just tested in Nike Settles Trademark Dispute with The Shoe Surgeon? These are board-level questions, not filing-docket questions — the theme of Your IP Strategy Is Only as Good as Your Board’s Questions. The consumer is being re-intermediated. Companies whose competitive advantage rests on owned, enforceable, extensible IP will ride that shift; those relying on channel position will feel it first.

For IP leaders, that is the opening. The report shows respondents overwhelmingly value patent and competitive intelligence earliest — at ideation and feasibility, before sunk cost makes a bad project politically impossible to stop. Yet most teams still use IP intelligence reactively: monitoring competitors and searching prior art, not shaping which bets get funded.

The strategic takeaway is simple. Treat freedom-to-operate, white-space mapping and competitive patent signals as a front-end input to portfolio decisions, not a back-end clearance check. Build the discipline of asking “what does the IP landscape tell us?” before commitment, and you convert a cost centre into a filter that protects budget, sharpens go/no-go calls, and compounds into an advantage competitors can’t easily copy.

The strategic signal for IP-intensive businesses isn’t the headline number. It’s buried in a quieter line: first movers can “secure intellectual property to build defensible ownership of key quantum computing applications.” Patent filings in quantum are already concentrating among a handful of well-capitalised leaders. In an emerging technology, the window to protect an invention closes from the outside in — every month you wait, more of the foundational landscape is claimed by someone else, and your eventual freedom to operate narrows whether or not you ever filed a thing.

Here’s the discipline the report rewards, and it’s the same pattern across every hot technology, not just quantum. The advantage doesn’t come from owning the technology — most players will eventually access it through cloud and quantum-as-a-service platforms anyway. It comes from a deliberate patent strategy: identifying the few use cases where your proprietary data, workflows, and applications create something competitors can’t easily replicate, and securing that first-mover advantage before valuations and the patent thicket make entry expensive. So the practical moves are unglamorous but decisive — map your exposure (including Q-Day cryptographic risk if you’re in financial services), pick the two or three hybrid use cases where IP could compound, get ownership and inventorship right from the first filing, and decide consciously where a patent beats a trade secret in a field that may be easy to design around. The mistake to avoid is treating quantum as a breakthrough to wait for rather than a capability — and an IP position — to build now. The companies that move will shape the standards; the rest will license them.

For more on why disciplined IP strategy beats the technology itself as a source of durable advantage, see AI Isn’t Your Advantage—Your IP Strategy Is (same lesson, applied to AI: in a hot emerging technology the moat is your protected assets, not the tech). On the principle that the decisive IP advantage is built years before it’s needed, see Built Before the Fight: What May’s IP Decisions Reward (directly on capturing position early rather than reacting). And on protecting IP before the window closes when moving into a new market, see Preparing for a New Launch (on securing rights early and managing the risk of something new reaching the market).

For IP-intensive businesses, this has a sharp edge, because everything AI accelerates is available to your competitors at the same subscription price. Speed gains get competed away; what remains is whatever you own that others cannot copy — your patents, brands, trade secrets and data. So the purpose-driven question De Cremer urges becomes, in IP terms: is AI helping you file faster, or helping you build a more defensible position? A firm using AI to churn out more of the same deliverables gains months; a firm using it to map competitor white space, strengthen invention capture and free its people for judgment work gains a moat. This is the pattern we examined in AI Isn’t Your Advantage—Your IP Strategy Is, and it starts with leadership discipline — Your IP Strategy Is Only as Good as Your Board’s Questions — because rushed deployment carries real risk, as we flagged in Your AI Agent Won’t Keep a Secret. The takeaway: before your next AI initiative, ask what proprietary, protectable value it will leave behind in two years. If the honest answer is “none — just speed,” you’ve found the urgency trap.