Part 4: We Found a Problem Patent. Now What?

The Freedom to Operate (FTO) Deep Dive: Questions Your Company Should Be Asking

This is Part 4 of a six-part series on freedom-to-operate analysis.  The series walks through what an FTO is, why it matters, how it works, what to do with the results, how it changes for software, AI, and international products, and the most common ways it goes wrong.  See the full table of contents here. 

Part 4: We Found a Problem Patent.  Now What?

An FTO that turns up a problem patent isn’t necessarily bad news.  It’s the FTO doing its job.  The bad outcome is finding the problem patent during litigation.  Finding it during an FTO is the whole point.  The harder question is what to do about it.  There are more options than most people realize, and the right choice depends on the strength of your invalidity case, how readily you can change the product, who owns the patent, and what stage of conflict you’re in.  This post walks through the main options.

What Are Our Options?

Design Around.  Sometimes a small engineering change moots the entire problem.  Modify the product to avoid one or more claim limitations.  This is best when the claim limitation is narrow, the change won’t ruin the product, and there are no pending continuations of the problem patent.  Make sure the design-around works under the doctrine of equivalents (covered in Part 3). A real design-around removes a limitation. Substituting one part for another rarely does the job.

• Bad design-around: “the claim requires a fastener; we’ll use a different fastener.” Both fasteners do the same job, so under DoE, you probably still infringe.
• Good design-around: “the claim requires a separate authentication step before the data transfer; we’ll restructure so authentication happens as part of the data transfer protocol itself, not as a separate step.” Now the architecture is genuinely different, and there’s no “equivalent” of a separate step in your product.
• Better design-around: target a limitation that was added by amendment during prosecution. The patent owner already gave up scope there.  They’re estopped from arguing equivalents.
• Best design-around: combine a structural change with an invalidity backup. Even if the design-around fails, the patent might be invalid; even if the patent is valid, the design-around works.  Layered defenses are stronger than single-point defenses.

License.  Negotiate a license.  This could be a per-unit royalty, a lump sum, or a cross-license.  Your negotiation leverage generally depends on how strong your invalidity case is and how readily you could design around.  It also depends on who owns the patent: an operating competitor wants to slow you down; a patent assertion entity wants money.

Challenge Validity.  There are several ways to attack a patent’s validity (which are generally outside the scope of this article), but the choice depends heavily on timing and the current state of the law.

Inter Partes Review (IPR).  Filed at the Patent Trial and Appeal Board (PTAB) and limited to invalidity arguments based on prior patents and printed publications.  Once the workhorse of patent challenges, with institution rates running around 65-70% for years, they have recently fallen somewhat out of favor.  In 2025, institution rates fell to a five-year low of 50%, with monthly rates dropping below 30% by late 2025.  First-quarter 2026 IPR filings were down roughly 66% year over year.  This decline comes from USPTO leadership changes that expanded discretionary denials.  IPRs are still available and still useful, but they are noticeably harder to get instituted than they were two years ago, and the current strategy increasingly turns on whether the petition can avoid discretionary denial.  Cost is typically in the $400,000 to $700,000 range through final written decision, more if appealed.

Post-Grant Review (PGR).  Broader than IPR (you can challenge on more grounds, including § 101 and § 112), but only available in the first 9 months after the patent issued.  These are rare in practice because the time window is so narrow.  The cost is similar to IPR.

Ex Parte Reexamination (EPR).  The patent owner and challenger don’t litigate against each other; the challenger submits prior art, and the USPTO decides whether to reexamine.  EPR is cheaper than IPR (often in the $50,000 to $100,000 range), but the challenger has very limited participation after the initial petition, and the patent owner gets to respond to the Examiner directly.  EPRs have recently surged (in early 2026) as petitioners pivoted away from IPR.

District Court Litigation.  Invalidity is always available as a defense (and often a counterclaim) in litigation.  The problem with relying on litigation invalidity is that you only get there by being sued (or by filing a declaratory judgment action, which has its own jurisdictional hurdles and is, in practice, very similar to being sued), and you have to take the case all the way to a judgment to get a binding ruling.  That can mean two to four years and several million dollars, which doesn’t really solve the problem you set out to solve with the FTO in the first place.  The FTO’s purpose was to determine whether you can sell the product safely.  “We’ll prevail on invalidity at trial” is a litigation strategy, not a clearance position.

Invalidity is a useful tool when you have strong prior art and a path to use it pre-litigation (IPR being the main option).  Invalidity is much weaker as a strategy when your only realistic forum is district court litigation, because by then you’re already in the situation the FTO was supposed to help you avoid.

Buy the Patent.  Buying the problem patent can be a clean solution, but it is usually very fact-specific.  This generally works best when the patent owner isn’t actively selling their own competing product.

Defensive Aggregator.  Organizations like RPX, AST, and Unified Patents acquire patents and license them to members or challenge them at the PTAB.  Worth considering in fields with heavy patent assertion entity activity.

Accept the Risk.  There are situations where this makes sense, but it should be well thought-out, with clear documentation of the rationale, sized to the company’s risk tolerance, and with a litigation reserve.

Covenant Not to Sue.  This is a contractual alternative to a license that can be useful when the patent owner has business reasons to avoid a formal license.

How Do We Choose Among These Options?

Here are the factors that usually drive the decision:

Strength of the Invalidity Case.  If you have strong prior art that wasn’t considered during prosecution, an IPR/EPR (or the threat of one) can help the negotiation.  If you don’t, invalidity isn’t a real option, and the conversation is about design-around or licensing.

Cost/Feasibility of the Design-Around.  Sometimes a small engineering change moots the problem entirely.  Sometimes the limitation is so central to how the product works that designing around it would mean rebuilding the product.  The cost of the change has to be weighed against the cost of every other option.

Who Owns the Patent?  An operating competitor wants to slow you down or extract a settlement that hurts you.  A patent assertion entity wants money and usually doesn’t care about anything else.  A defensive aggregator wants leverage.  The right response depends on what the patent owner actually wants.

Stage of Conflict.  Pre-assertion, you have all options available.  Once you’re sued, IPR is still possible (within a one-year window), but design-around is harder to deploy without admitting infringement, and licensing leverage shifts to the patent owner.  The earlier you find the problem, the more options you have.

Litigation Budget and Risk Tolerance.  IPR runs several hundred thousand dollars.  District court litigation runs several million.  A startup may not have the cash for full litigation; a public company may not have the risk tolerance for ongoing uncertainty.  Real-world budget constraints often determine which legal tools are actually available.

Strategic Relationship with the Patent Owner.  If the patent owner is also a current customer, a potential acquirer, or a partner, the situation is trickier.  Picking a fight with a strategic counterparty over a patent dispute can cost more than the patent itself.

In practice, the right answer usually combines options.  A design-around plus an invalidity backup.  A license negotiated under the threat of an IPR.  An acceptance of risk for low-likelihood threats combined with a design-around plan ready to deploy if assertion happens.  The single-option responses are rarely the strongest ones.

No single response works for every problem patent.  The right approach usually combines two or three of these options (a design-around backed by an invalidity argument, or a license negotiated under the threat of an IPR), not a single silver bullet.  The starting point is to understand the actual options and what each one does well.  The cheap and easy responses are available when the problem is found early.  They aren’t available once you’re in litigation.

Part 5 picks up special situations: how FTO works differently for software and AI, what changes for standards-essential patents, and what international manufacturing or sales mean for the analysis.