Drafting mistakes that cause litigation headaches later

What happened

Alice held patents on mitigating settlement risk using a computer as an intermediary. The Supreme Court applied a two-step framework: first, is the claim directed to a patent-ineligible concept such as an abstract idea? Second, if so, do the claim elements — individually and as an ordered combination — add an "inventive concept" that transforms it into something more?

The Court held the claims were directed to the abstract idea of intermediated settlement, and that reciting a generic computer performing generic functions added nothing inventive. All claims were held ineligible.

Why it's a drafting problem

The decision hit an enormous number of already-issued software patents, and the ones that fared best afterward were those drafted as a specific technical improvement rather than as a business or economic practice implemented on a computer. In Enfish v. Microsoft (Fed. Cir. 2016), claims to a self-referential database table survived at step one precisely because they were framed as an improvement to how a computer functions.

Often the same underlying invention could have been framed either way. The framing was a drafting decision made years before anyone had heard of Alice.

What happened

Biosig's patent covered a heart-rate monitor with electrodes in a "spaced relationship" with each other. The dispute was whether that phrase told a skilled reader what was actually claimed.

The Supreme Court rejected the Federal Circuit's forgiving "insolubly ambiguous" test and replaced it with a stricter one: a claim is indefinite if, read in light of the specification and prosecution history, it fails to inform those skilled in the art about the scope of the invention with reasonable certainty.

Why it's a drafting problem

Relative and relational terms are everywhere in patent claims — "substantially," "about," "proximate," "spaced relationship," "sufficient." They're convenient during drafting because they feel broad. Under Nautilus they're a liability whenever the specification gives no objective way to determine what falls inside.

The fix costs almost nothing at drafting time: a sentence in the spec defining the term, giving a range, or describing how a skilled person would measure it. It's unfixable after filing, because you cannot add new matter.

What happened

For years, drafters relied on a simple rule: avoid the word "means" and § 112(f) wouldn't apply, so your functional language wouldn't be limited to the structures disclosed in the spec. The Federal Circuit, sitting en banc on this issue, overruled the "strong" presumption that had made that rule reliable.

The claim term was a "distributed learning control module." The court held it was a means-plus-function term despite never using the word "means," because it recited function without sufficiently definite structure. The specification failed to disclose corresponding structure — an algorithm — so the claims were held indefinite.

Why it's a drafting problem

"Module," "unit," "mechanism," "element," "device," "component" — these are nonce words. Paired with a function and no structural detail, they invite a court to treat the claim as means-plus-function and confine it to whatever structure the spec disclosed. If the spec disclosed none, the claim can be invalid outright.

This is especially dangerous in software, where it's tempting to describe what a component does and skip how it does it.

What happened

Festo narrowed claims during prosecution to secure allowance. Later it sued a competitor whose product didn't literally infringe but was arguably equivalent.

The Supreme Court held that a narrowing amendment made for reasons of patentability triggers prosecution history estoppel — but rejected the Federal Circuit's absolute bar. Instead there's a rebuttable presumption that the patentee surrendered everything between the original and amended claim. The patentee can rebut it by showing the equivalent was unforeseeable at the time, that the amendment's rationale was only tangentially related, or some other comparable reason.

Why it's a drafting problem

Amendments feel free in the moment. The examiner objects, you narrow, the case advances. But you may have permanently surrendered the design-around space a competitor will later occupy — and you won't find out until you try to enforce.

The alternatives are real and routinely underused: argue without amending, amend along a different axis that doesn't give up commercially relevant ground, or accept the narrower claim here while pursuing the broader one in a continuation.

What happened

The Federal Circuit settled how claim terms are to be construed. Intrinsic evidence governs — the claims themselves, the specification, and the prosecution history — with the specification described as the single best guide to the meaning of a disputed claim term. Dictionaries and expert testimony are secondary.

Why it's a drafting problem

This is the case that makes everything else on this page matter. Whatever you wrote in your specification is the primary evidence a court will use to decide what your claims mean. You are writing the dictionary that will be used against you — or for you — years later, by someone with every incentive to read it narrowly.

It cuts both ways. A specification that carefully defines its terms and describes multiple embodiments supports the construction you want. One that describes a single embodiment in absolute language ("the invention requires…") hands the other side an argument that your claims are limited to exactly that.

What happened

Amgen claimed an entire genus of antibodies defined by what they do — bind to specific residues and block a receptor — rather than by their structure. The specification described some examples plus a research method for finding others.

The Supreme Court unanimously held the claims invalid for lack of enablement. The specification has to enable the full scope of what's claimed, and offering a roadmap that requires substantial trial and error to reach the rest of the genus isn't enough.

Why it's a drafting problem

Broad functional claiming is seductive: define the invention by the result it achieves and you seem to capture every way of achieving it. Amgen is the price. Claim breadth has to be matched by disclosure breadth, and the mismatch is invisible at allowance but fatal at trial.

Although Amgen arose in biotech, the reasoning isn't limited to it — any broad functional claim in any field raises the same question of whether the specification actually taught the full scope.