On the insurability of risks around intellectual property
Patents, trademarks, copyrights and closely guarded trade secretsare the real capital of many companies today. At technology, pharmaceutical or software firms, most of the corporate value is not sitting in machines or buildings anymore and moved into ideas. So, if intellectual property is this valuable, why not simply insure it, the way you insure a building against fireor a car against a crash? The surprising answer is: this does only partly work. And of all things, the most valuable parts are the hardest to insure.
One big term for many different risks
“IP risk” sounds like a single thing, but it is really a whole bouquet of very different aspects. A company can accidentally infringe someone else's IP, e.g. a patent, and be sued. Conversely, it can watch its own patent be challenged by a competitor and declared invalid. Enforcing your own rights in court can become expensive. And finally, a patent or trademark can simply lose economic value because a new technology overtakes it or the market shifts. This distinction is the key to the entire topic, because whether a risk is insurable does not depend on how valuable the IP is – it depends on what kind of risk we are dealing with.
To understand why this is the case, it helps to remember what an insurance policy actually needs. Insurance works on a simple idea: many people pay a manageable premium, and when a loss occurs, the few who are affected are compensated out of that pool. For that to add up, the insurer needs three things:
First, the loss must be random – no one may already know at the time the contract is signed that it will occur. Second, it must be reasonably calculable, otherwise no fair premium can be set. And third, there needs to be a sufficient number of comparable cases from which a reliable benchmark can bebuilt.
And this is where it gets unfavourable: insurance traditionally protects against losses – not against forgone profit. You can insure your car against a crash, but not the hope that it will rise in value. And the greatest value in intellectual property consists precisely of such hopes: future market share, licensing income, technological advantage.
The paradox: the more valuable, the harder to insure
The cost of a lawsuit can be estimated reasonably well – lawyers, experts, court fees. Such litigation-cost risks are therefore genuinely insurable, and that is what most classic IP insurance products revolve around. But how much revenue a patent will generate over the next ten years, whether an invention will catch on in the market or be overtaken by a competitor – no one can seriously calculate that. These future opportunities are the very heart of the economic value – and for exactly that reason they can hardly be covered. That is the real paradox: the insurable components are often the less valuable ones, and the most valuable ones are barely insurable. It is never the patent itself that gets insured, but always only a specific, clearly defined danger connectedto it.
This also explains why there is no single “IP insurance,” but rather many specialised products that each cover one piece. Some take on the cost of legal defence, others the enforcement of your own rights, and still others – such as the W&I policies common in M&A – cover the seller's warranties about its intellectual property when a company is bought. There are litigation policies built specifically for the exceptional cost of patent disputes, contingent-risk products for a known legal question whose financial outcome is still open, and D&O cover that steps in when a company's directors are held personally liable for an IP-related decision. Each one carves out a slice that can be defined, quantified and priced, and leaves the rest untouched. This looks fragmented, but it is not a weakness of the market: it is a smart response to the nature of these risks. It also explains a common misunderstanding – that the modest size of the classic IP-insurance market means the industry has failed to grasp how important intellectual property has become. The opposite is true. The market is small not because demand is missing, but because the pieces that are genuinely calculable are only a part of the whole. What gets insured is what can sensibly be calculated.
And what about artificial intelligence?
AI sharpens this picture rather than changing it. Anyone who trains AI systems or commercialises their output is still on legally uncertain ground: court decisions, empirical values and clear rules are all missing. For insurers, that means risks that are barely calculable – and if a single mistake hits many users at once, losses can even threaten on a large scale. At the same time, the very same technology could help assess risks better in the future by analysing vast amounts of IP and court data. The limits of insurability may shift as a result – but they will not disappear. The underlying logic stays the same: whether an IP risk can be insured does not depend on how much the right is worth, but on how well the specific danger can be calculated. What gets insured is not patents or trademarks as such, but the individual, tangible risks attached to them. The rest – the great entrepreneurial bet on the future– remains what it always was: a matter for the entrepreneur, not an insurer.
