July was a month about ownership. Not compliance, not litigation skill – ownership.
Google’s edge sits on TPU technology that competitors must come back and license. Sanofi lost the mRNA product race and is now taxing the winners through delivery patents it happens to hold. A Melbourne pool servicing company paid $60,000 for skipping the one search that would have told it someone else already owned the name it wanted.
Twelve posts published this month split cleanly into four groups worth reading as a set: who owns the platform underneath the product, who owns the brand, what a dispute actually costs the winner as well as the loser, and whether strategy execution matches what the strategy document says. Read together, they make the same point from four directions. IP strategy is not a filing exercise. It is a decision about who gets to charge whom.
Owning the platform: AI, compute and the assets everyone else needs
The Cash Behind the Compute makes the case plainly. Google’s advantage is not its ad engine or its cash pile – it is the TPU technology competitors must license. Owning the foundation, not outspending the field, is what compounds.
The same pattern showed up in a messier form in trade secrets. Autonomous AI agents are now capable of exfiltrating data without an identifiable human defendant, which breaks the basic assumption trade secret law depends on. The practical response is to lean harder on patents for the assets that matter most, and to tighten data governance around what agents can see and move.
On the licensing side, a week of AI copyright suits and platform-patent claims showed where the money actually flows. Sanofi is asserting lipid nanoparticle delivery patents against both Pfizer and Moderna. It lost the product race but still owns part of the platform, and a royalty claim does not need an injunction to matter.
Brand as owned asset: trade marks under test
Costco’s Kirkland strategy is the positive version of the same lesson. A $90 billion private-label business was built by registering trade marks before engaging manufacturers, and by locking IP ownership into the supply contracts themselves. Registered rights plus contract discipline turn a commodity into an owned asset.
Ownership has to survive scrutiny once it exists. In the Full Federal Court’s ruling on the Registrar’s revocation powers, a charity kept its BROWN NOSE DAY registration despite a challenge from the RED NOSE DAY owner. Registrations get safer the longer they survive unchallenged, and a mark’s own meaning is its strongest defence.
The cost of getting this wrong is concrete, not theoretical. A $60,000 trade mark judgment against a pool servicing company turned on a basic gap: an ASIC business name search is not a trade mark search, and business name registration gives no right to use the name at all.
The cost of disputes: who pays and who collects
Three 2026 cases – Anthropic’s copyright settlement, Disney’s UPC injunction, and Apple’s Masimo verdict – show what happens when input sourcing, data provenance and freedom-to-operate reviews are skipped before launch. IP strategy decides which side of the ledger a business ends up on.
Winning is not the same as collecting cleanly. A mixed litigation result still leaves the largely successful party covering part of its own costs, which argues for narrowing legal grounds before filing, and for never sending a cease-and-desist letter without provable ownership behind it.
Preparation decides more than the merits do. A week of IP disputes made the point that leverage comes from strategic infrastructure – enforcement venues aligned to where a counterparty actually earns its revenue, documented data provenance, and patents stress-tested for validity before they are asserted, not after.
Execution is the strategy: ownership, capital and stress-testing
AI transformation failures are usually ownership failures, not design failures. An IP strategy imposed top-down, without genuine engagement from the teams who have to act on it, stays a document rather than becoming a portfolio.
Capital allocation runs into the same wall without a clear centre. Every IP investment becomes a contested debate when a company lacks a strategic centre to allocate against – coherence, not breadth, is what compounds value over time.
The stress test is worth running before the world forces it on you. Wargaming an IP strategy against geopolitical disruption exposes weaknesses in filings, licensing structures and freedom-to-operate positions while they are still cheap to fix.
What to watch
Put the four threads together and July’s message is consistent. Ownership of the platform, the brand, the enforceable right and the internal mandate to act on strategy is what decides who pays and who gets paid. None of it happens by default. It is filed, registered, documented and mobilised, or it is not there when the bill arrives.
The useful question for your next strategy review is not whether you have an IP strategy. It is whether, item by item, you actually own the things that strategy assumes you own – and whether the people who have to act on it would say the same.