Week 9 - Enforcing Patent Rights (II)

Doctrine of Equivalents Limits

Equivalents Cannot Cover New, Nonobvious Inventions The doctrine of equivalents should not let a patent owner claim someone else's later invention that is new and not obvious. In this view, a product cannot be both "insubstantially different" from the patent and nonobvious at the same time, and the text points readers to a law review article on this link.

Four Limits on the Doctrine of Equivalents The text names four limits on the doctrine of equivalents: prosecution history estoppel, the disclosure-dedication rule, the all-limitations and specific exclusion rule, and ensnarement. It explains that the next four subsections cover each of these limits in turn.

Prosecution History Estoppel

The Basic Idea of Prosecution History Estoppel Prosecution history estoppel stops a patent owner from using a lawsuit to win back claim coverage it gave up while applying for the patent. For example, an inventor who narrowed an electrode-distance range from 3–10 centimeters to 3–5 centimeters to get past the examiner cannot later sue a competitor whose device uses 7 centimeters.

Festo Extends Estoppel to Foreseeable Equivalents Under the Supreme Court's Festo decision, a patent owner who narrows a claim by amendment is presumed to have given up everything between the original claim and the amended claim. The owner can overcome that presumption only by showing the equivalent was unforeseeable, was only loosely related to the reason for the amendment, or could not reasonably have been described for some other reason.

Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd. (Festo VIII) 535 U.S. 722 (2002) Background

Festo's Patented Device Festo owns two patents on an improved magnetic rodless cylinder, a piston-driven device that uses magnets to move objects along a conveyor system. The device has many industrial uses, from sewing equipment to the Thunder Mountain ride at Disney World.

How Festo's Patents Were Amended and How SMC's Device Differs Both of Festo's patents were amended during the application process to add a pair of one-way sealing rings, and the Stoll Patent also added a requirement that the sleeve be made of magnetizable material. SMC's competing device uses a single two-way sealing ring and a nonmagnetizable sleeve, so it does not literally infringe, but Festo claimed it infringes as an equivalent.

Procedural History

The Federal Circuit's Ruling and the Questions Presented When an applicant narrows a claim after the Patent Office rejects it, the applicant cannot later argue that the broader, abandoned version was merely an "equivalent," and the Federal Circuit held that such narrowing gives up all equivalents to the amended element. Festo argued that this ruling went too far in two ways, by applying to every kind of amendment and by fully barring all equivalents, and the Supreme Court agreed to hear the case.

SMC's Estoppel Argument and the Case's Path Through the Courts SMC argued that Festo gave up these very features when it narrowed its claims to get its patents, but the district court disagreed and a Federal Circuit panel affirmed. After the Supreme Court sent the case back, the full Federal Circuit reversed, holding that any narrowing amendment made to comply with the Patent Act creates estoppel and that estoppel completely bars equivalents for the amended element.

The Federal Circuit Replaces the Flexible Bar With a Complete Bar The Federal Circuit admitted that its earlier cases treated estoppel as a "flexible bar" that blocked some equivalents but not others, depending on the amendment. It overruled those cases, calling the case-by-case approach unworkable and arguing that a complete bar would make infringement outcomes more certain.

Legal Standard

Balancing Protection Against Copying With Clear Boundaries The Court explains that the doctrine of equivalents protects patent owners from copycats who make only minor changes, but it also makes it unclear where a patent's coverage ends. To reduce that uncertainty, competitors may rely on the prosecution history, which is the public record of what happened at the Patent Office.

Patents as Property With Clear Boundaries Patents reward innovation with a temporary monopoly, which is a property right whose boundaries should be clear so that owners and the public both know what is covered. For that reason, the law requires inventors to describe their inventions in exact terms, yet language can never fully capture an invention.

Why Patents Cover More Than Their Literal Words Because words often lag behind new inventions, reading patents only by their literal terms would let copycats avoid infringement with trivial substitutions. So a patent covers both its literal claims and their equivalents, even though this makes patent scope less certain.

The Cost of Uncertainty and the Doctrine's Long History Uncertainty about a patent's reach can scare competitors away from lawful products, lead them to invest by mistake in infringing ones, and produce wasteful lawsuits. Even so, the Court has repeatedly accepted this uncertainty as the price of protecting inventors, as in Winans v. Denmead and Graver Tank, despite dissents calling for stricter rules.

Dissenting Concerns and Deference to Congress In Graver Tank, Justice Black warned in dissent that competitors would have to guess, at the risk of heavy damages, how far a court might stretch a patent's language. In Warner-Jenkinson, however, a unanimous Court reaffirmed the doctrine of equivalents and said that only Congress, not the courts, should get rid of it.

Estoppel as a Rule for Reading Patent Claims Prosecution history estoppel requires courts to read patent claims in light of what happened at the Patent Office, including claims that were canceled or rejected. The doctrine of equivalents, by contrast, lets patent owners cover minor variations that they did not capture when drafting the original claim.

Narrowing After Rejection Is a Concession If an applicant first claimed something broader and then narrowed the claim after a rejection, the applicant cannot later argue that the surrendered material was an unforeseen equivalent. Choosing to amend rather than appeal is treated as admitting that the patent does not reach as far as the original claim.

Protecting the Patent Office's Gatekeeping Role Without estoppel, inventors could sidestep the Patent Office by winning back in court the very coverage they gave up to get the patent. Because the doctrine of equivalents rests on the idea that words cannot fully capture an invention, it makes little sense when the inventor clearly knew the words for both the broader and narrower claim and chose the narrower one.

Which Amendments Trigger Estoppel The first question is what kinds of amendments create estoppel, and Festo argued that only amendments narrowing the invention's substance, such as those avoiding earlier inventions, should count. The Court noted that Warner-Jenkinson applied estoppel only to amendments made for a limited set of reasons, such as avoiding prior art or addressing concerns like obviousness.

Estoppel Applies to Any Patent Act Requirement Although estoppel usually comes up when an amendment avoids earlier inventions, the Court held that a narrowing amendment made to satisfy any Patent Act requirement can create estoppel. Those requirements include being useful, new, and nonobvious, as well as properly describing and enabling the invention.

Legal Reasoning

Section 112 Amendments—Cosmetic Versus Narrowing Festo argued that amendments to meet Section 112's description rules only concern the application's form, such as clarifying a term, and should not cause estoppel. The Court answered that what matters is the amendment's effect: a purely cosmetic change does not trigger estoppel, but a necessary change that narrows the patent's scope may.

The Court Rejects the Complete Bar The second question is whether estoppel blocks all equivalents to a narrowed element, and the Court disagreed with the Federal Circuit's complete-bar rule. Even after an amendment, language remains imperfect, so a narrowing amendment should not give up equivalents that were unforeseeable or only loosely related to the reason for the amendment.

Precedent and Inventors' Settled Expectations An amendment does not mean the inventor suddenly drafted perfectly, and the Court's past cases have always applied estoppel flexibly rather than rigidly. The Federal Circuit also ignored Warner-Jenkinson's warning that courts should not upset inventors' settled expectations, since changing these rules is Congress's job.

Fairness to Past Inventors and Who Bears the Burden Inventors who amended under the old rules had no reason to think they were giving up all equivalents, so applying a harsher rule to them would be unfair. Under Warner-Jenkinson, if the patent owner cannot explain an amendment, courts presume it was made for patentability reasons and that everything between the broader and narrower language was surrendered.

The Patentee Must Show the Equivalent Was Not Surrendered The Court held that the patent owner must prove that an amendment did not give up the particular equivalent at issue, because a narrowing amendment is presumed to give up the ground between the original and amended claims. The owner can overcome this presumption by showing the equivalent was unforeseeable, was only loosely related to the amendment's purpose, or could not reasonably have been described for some other reason.

A Rebuttable Presumption, Not a Complete Bar This presumption is not just the complete bar under another name, because patent owners can still rebut it. To do so, the owner must show that, at the time of the amendment, a skilled person could not reasonably have written a claim literally covering the alleged equivalent.

The Outcome—Vacated and Sent Back Festo had not yet overcome the presumption, because it admitted the sealing-ring and sleeve amendments were made for patentability reasons, so the real question is what those amendments gave up. The Court vacated the Federal Circuit's judgment and sent the case back for lower courts to decide that question, noting that SMC might well win.

Festo Flow Chart

Johnson & Johnston Assocs., Inc. v. R.E. Service Co., Inc. 285 F.3d 1046 (Fed. Cir. 2002) Background

The Case Begins Johnston sued RES for infringing its '050 patent, and a jury found willful infringement under the doctrine of equivalents and awarded about $1.14 million. The full Federal Circuit reheard the case and reversed, holding that RES could not have infringed under the doctrine of equivalents.

The Patented Invention The '050 patent concerns making printed circuit boards, where workers handling thin copper foil by hand can damage or contaminate it. The invention attaches the fragile foil to a stiffer aluminum sheet so workers can handle it safely, and the aluminum can be removed and recycled afterward.

What the Patent Description Says A band of flexible glue joins the foil and the backing sheet at the edges, which makes the product stiffer, easier to handle, and less likely to be ruined. The description also says that aluminum is preferred but that other metals, such as stainless steel or nickel alloys, and sometimes polypropylene, may be used.

The Patent Claims at Issue Claim 1 covers a copper foil sheet joined at its edges by flexible adhesive to "a sheet of aluminum" that is later thrown away. Claim 2 covers a similar product with copper foil attached to both sides of the aluminum sheet.

Procedural History

RES's Steel Product and the Trial Court's Rulings In 1997, RES began making circuit-board products that used steel instead of aluminum as the backing sheet, and Johnston sued. The trial court found no literal infringement and rejected RES's argument that Johnston had given steel to the public by describing it without claiming it, so the equivalents question went to trial.

The Appeal and the Maxwell Case On appeal, the parties disagreed over whether the Maxwell case or the YBM Magnex case should control. In Maxwell, the court held that a shoe-fastening method described in the patent but not claimed belonged to the public, so using it could not infringe as an equivalent.

Legal Standard

How YBM Magnex Narrowed Maxwell Maxwell reasoned that not claiming the alternatives kept the Patent Office from reviewing them, so a skilled reader would see them as given to the public. YBM Magnex later limited Maxwell to cases where the unclaimed alternative is clearly separate from the claimed invention, so the court now had to decide whether equivalents can ever cover described but unclaimed material.

The Claims Define the Patent Both the Supreme Court and the Federal Circuit hold that the claims alone set the limits of a patent and give notice to the Patent Office and to competitors. Infringement is decided by comparing the accused product with the claims, not with examples in the patent's description.

Early Cases Say What Isn't Claimed Is Public In the 1880s, the Supreme Court said in Miller and Mahn that failing to claim something shown in a patent gives it to the public. The Court explained that the public has a clear right to use whatever is not specifically claimed.

Graver Tank and the Modern Doctrine of Equivalents The doctrine of equivalents stretches a patent beyond its literal words, and Graver Tank II said equivalence depends on the patent, earlier inventions, and the facts of the case. In the earlier Graver I decision, broad claims to "silicates" were struck down, and the accused product swapped in manganese silicate for the "alkaline earth metals" named in the narrower claims that survived.

Legal Reasoning

Why Graver Tank Was Not a Dedication Case The Court found infringement because the change was minor and the product worked in substantially the same way to get the same result. The patent owner in that case had not given anything to the public, because it had actually claimed the "equivalent" material, even though those claims were later held too broad.

Warner-Jenkinson Is Also Not a Dedication Case In Warner-Jenkinson, the Supreme Court refused to abolish the doctrine of equivalents and held that an added pH limit of 6.0 did not automatically rule out equivalents for that element. That patent never described running the process at a pH of 5.0, so it did not involve giving described material to the public.

Why the Maxwell Rule Makes Sense Describing something without claiming it gives it to the public, and using equivalents to win it back would undercut the principle that claims define the patent. Otherwise, applicants could write broad descriptions with narrow claims to avoid a close look by the Patent Office and then use equivalents to get broader coverage later.

The Court's Holding Johnston's claims were limited to aluminum, but its description mentioned steel, so Johnston cannot use the doctrine of equivalents to stretch its claims to cover steel. To the extent YBM Magnex says otherwise, the full court overruled it.

Remedies for Owners Who Forget to Claim Something A patent owner who accidentally fails to claim described material can file a reissue application within two years to try to broaden the claims. The owner can also file a continuation application claiming that material, which Johnston in fact did twice.

Dissent on why Disclosure rules are too restrictive

Judge Newman's Dissent—Drawing Lines Judge Newman argued that patent owners often have to draw lines to describe their inventions precisely, and that a rigid rule that punishes disclosure is too simple and conflicts with patent policy. She stressed that sharing technical information is a core purpose of patents, and punishing it makes a patent less useful as a source of knowledge.

The Dissent's Concern About Copiers Under the majority's rule, a copier can avoid even an equivalents claim simply by stepping across a line the patent owner drew in a described range. She cited a 1994 survey in which 65% of R&D managers named how easy it is to get around patent claims as the main reason not to invest in patents, and 47% worried about disclosing technical information without enough protection.

The Dissent on Innovation and Fairness Judge Newman said that inventors who take risks drive economic growth, so courts should not shift the balance between inventors and imitators without weighing the effects. In her view, a neat bright-line rule is not necessarily good policy, and fairness, the foundation of due process, should come before such rules.

Disclosure-Dedication Rule

A legal principle that holds that if an inventor publicly discloses an unclaimed invention in a patent, that invention is dedicated to the public. This rule ensures that applicants cannot use broad descriptions to retain undisclosed rights while limiting the scope of their claims.

The All-Limitations Rule

The all-limitations rule says that every requirement in a patent claim matters and cannot be ignored or made meaningless. So to prove infringement by equivalents, each requirement must have an equivalent in the accused product, rather than comparing the invention as a whole.

The Specific Exclusion Rule

A related rule says that equivalents cannot be used to cover something a claim specifically excludes, because the owner implicitly gave that material up. The SciMed case is offered as the main example of both rules.

SciMed v. ACS Background

The Case Begins SciMed sued ACS for infringing three patents on balloon catheters, and the trial court ruled for ACS on summary judgment. The trial court found that the claims required a structure ACS's devices lacked and that there was no infringement by equivalents either, and the appeals court agreed.

How the Catheter Is Built Balloon catheters are used in heart procedures to clear blocked arteries, and SciMed's catheters have a stiff first section, a flexible second section with a balloon, and a transition section between them. The transition section connects the two shaft sections and gradually changes the stiffness from one to the other.

he Catheter's Two Channels The catheter has two channels: one slides over a guide wire to steer the device to the blockage, and the other carries fluid to inflate the balloon. In SciMed's design, the guide wire enters near the front end of the catheter rather than at the surgeon's end, and the inflation channel runs the full length to the balloon.

Side-by-Side Versus One-Inside-the-Other Both sides agreed there are only two known ways to arrange the channels: side by side ("dual") or one inside the other ("coaxial"). ACS's devices use only the side-by-side design, while SciMed's preferred design uses the coaxial arrangement.

Claim 19 as the Sample Claim Claim 19 of the '594 patent stands in for all the asserted claims. It describes a catheter with a stiff metal first section, a more flexible second section, and a metal transition piece that gradually becomes less stiff, plus an inflation channel "separate from" the guide-wire channel.

Procedural History

The Trial Court Limits the Claims to Coaxial Designs Based on the patents' shared description, the trial court read the claims as covering only coaxial catheters, not side-by-side ones. It relied especially on language calling the coaxial structure the "basic sleeve structure for all embodiments of the present invention".

The Result Below and SciMed's Appeal Given that reading, SciMed admitted there was no literal infringement, and the court also found the two designs too different for a jury to find equivalence. SciMed appealed both the claim reading and the summary judgment.

The Key Question on Appeal The narrow question was whether the patents' shared description limits the claims to coaxial catheters, and the application record at the Patent Office said nothing helpful on this point. The appeals court agreed with the trial court that the description rejects the side-by-side design and limits the claims to coaxial catheters.

Legal Standard

Reading Claims in Light of the Description SciMed argued that the trial court made the classic mistake of adding a limit from the description into the claims. The appeals court disagreed, explaining that the trial court properly read the claims in light of the description, as the law requires.

The Watts Example When a patent's description makes clear that the invention does not include a feature, the claims do not cover that feature, even if the claim words alone seem broad enough. In Watts, the description described only one way to seal pipe joints, misaligned thread angles, so the claim was limited to that method.

The Toro Example In Toro, the question was whether a vacuum-blower's cover "including" a restriction ring had to be permanently attached to that ring. Because the description showed only an attached ring and praised that one-piece design, the court held that the claim required the ring to be attached

The Moore and Kodak Examples A structure falls outside the doctrine of equivalents when the claims clearly exclude it, whether openly or by implication, as in Moore, where adhesive covering a "minority" of a form's length could not be equivalent to a claim requiring a "majority". Similarly, in Kodak, a claim requiring "an inert gas atmosphere" excluded processes using heated air, so those processes could not infringe by equivalents.

The Sage Example In each of these cases, defining the claim to exclude certain material meant the owner gave it up, and Sage reached the same result for a syringe-disposal container with a specific slot design. The Sage court said the owner could have sought broader claims, so the owner, not the public, must bear the cost of not doing so.

Legal Reasoning

Applying These Cases to SciMed—The Abstract Like those cases, SciMed's description repeatedly shows that the invention uses coaxial channels, so the claim phrase "separate from" must mean coaxial. Right from the start, the patents' abstract describes the inflation channel as ring-shaped, which means coaxial rather than side by side.

Criticizing Earlier Side-by-Side Designs The patents criticize earlier catheters for using side-by-side channels, saying they were bulkier and stiffer than desired. By contrasting itself with that older design, SciMed showed that its claims should not be read to cover it.

The "Summary of the Invention" The patents' summary and conclusion repeatedly describe "the present invention" as having a ring-shaped inflation channel around an inner guide-wire tube. Describing the coaxial design as "the present invention" is strong evidence that the claims do not cover the opposite design.

The Most Telling Passage The most important passage calls the coaxial structure "the basic sleeve structure for all embodiments of the present invention". That language defines SciMed's invention in a way that excludes side-by-side channels.

Rejecting the "Preferred Version Only" Argument SciMed argued that the coaxial references described only its preferred version, not a limit on all claims. The court rejected this because the description repeatedly defines "the invention" as coaxial and criticizes side-by-side designs.

"All Embodiments" Means All The words "all embodiments of the present invention" are broad and clear, and it is hard to imagine how the patents could have been clearer that coaxial channels were required. Because SciMed knew about the side-by-side design and criticized it, this was a clear case of giving up that design.

No Literal Infringement Because the patents make clear that the claimed channels are coaxial, the trial court correctly read them as giving up the side-by-side design. SciMed admitted that under that reading there was no literal infringement, so summary judgment for ACS was proper.

No Infringement by Equivalents The court agreed that the doctrine of equivalents does not apply, because SciMed's patents identified, criticized, and gave up the side-by-side design. Having done that, SciMed cannot use equivalents to cover a structure it specifically excluded from its claims.

Linked to the Rule Against Ignoring Claim Terms This principle is like the rule that equivalents cannot be used in a way that wipes out a claim requirement, so a patent requiring a "non-metallic" device cannot be used against a metal one. Either way, the question is whether the patent clearly excludes the alleged equivalent, openly or by implication.

Applying the Rule to SciMed SciMed's patents openly recognized and rejected the side-by-side design as inferior, so allowing SciMed to recapture it would undermine the public notice that patents are supposed to give. SciMed could have drafted its patents to cover both designs but did the opposite, so no reasonable jury could find infringement by equivalents.

Ensnarement

The doctrine of equivalents cannot stretch a patent to cover technology that already existed before the invention (the "prior art"). This is because the Patent Office would never have granted a patent that broad, and accused infringers often argue they are just copying what already existed.

Wilson Sporting Goods Co. v. David Geoffrey & Associates 904 F.2d 677 (Fed. Cir. 1990) Background

The Competitors Wilson and Dunlop are major head-to-head rivals in the golf ball business, each selling well-known brands. Wilson accused Dunlop of infringing several specific patent claims and also made a general accusation against DGA, the distributor of Dunlop's Slazenger ball.

Why Dimples Matter Because golf rules tightly limit a ball's size and weight, designers have focused on surface dimples, which help the ball fly higher and farther. Even so, the science of dimples is complicated and imprecise, since dimples can vary in number, shape, depth, and placement.

The Patented Dimple Layout Wilson's patent is about where the dimples sit on the ball, not their shape, with the goal of spreading them more evenly. To do this, the ball's surface is divided into 20 imaginary triangles, which are then split into 80 smaller triangles.

The Six "Great Circles" The lines dividing the triangles form six circles around the widest part of the ball, and every claim requires that no dimple touch any of these circles. This layout gives the ball six lines of symmetry, compared to just one on earlier balls.

Getting the Patent Wilson's employee applied for the patent in 1984, and the Patent Office approved all 27 claims right away without comment. The patent was issued in 1985, and its only independent claim, claim 1, describes the dimple-free great circle design.

The Dependent Claims The other 26 claims build on claim 1 and add more detailed rules about how many dimples go in each triangle and where. For example, one claim requires the same number of dimples in each central triangle, while others place dimples on the edges of triangles.

The Earlier Uniroyal Ball The earlier technology included a Uniroyal golf ball from the 1970s that used the same 20-triangle, six-circle layout. On that ball, however, 30 or more dimples crossed the great circles by a small amount.

Dunlop's Accused Balls Dunlop's four balls use the same six-circle layout, but their circles are not dimple-free as the patent literally requires. Taking the evidence in the light most favorable to Wilson, about 60 dimples on each ball slightly cross the circles.

Legal Standard

Dunlop's Main Argument Wilson argued only that Dunlop infringed by making something "equivalent," not by copying the patent exactly. Dunlop responded that its balls are basically no different from the old Uniroyal ball, so covering them would also wrongly cover that older ball.

The Equivalence Test and Its Limit A product can infringe if it does basically the same job, in basically the same way, with basically the same result as the patented invention. Even then, there is no infringement if that range of equivalents would also cover the prior art, and the court decides this question itself.

Claims Are Not "Expanded" The court explained that it is inaccurate to say the doctrine of equivalents "broadens" a patent's claims, since the claims themselves stay the same. What actually expands is the patent owner's right to block products that are equivalent to what the claims describe.

Why Prior Art Sets the Limit Prior art limits equivalents because a patent owner should not win in court coverage that the Patent Office would never have granted. The doctrine exists to stop people from cheating a patent, not to give patent owners more than they could have lawfully obtained.

The "Hypothetical Claim" Test To make the analysis easier, the court imagines a claim broad enough to cover the accused product and asks whether the Patent Office would have approved it over the prior art. If that imagined claim would have been rejected, the patent owner cannot get that coverage in court, and if it would have been approved, the prior art is no barrier.

Wilson Bears the Burden The court dropped the "expanded claim" wording and noted that Wilson is seeking coverage beyond what the examiner ever considered. Wilson must therefore prove that its requested scope would not cover the Uniroyal ball, though its actual patent claims remain valid either way.

The Specific Question The question was whether Wilson could have patented an imagined claim broad enough to cover Dunlop's balls. That claim would have to allow 60 dimples to cross the great circles by up to nine thousandths of an inch.

Legal Reasoning

Comparison and Holding The court found that the imagined claim and the Uniroyal ball were only slightly different, since both had dimples crossing the circles in similar numbers and amounts. Because the imagined claim would have been obvious in light of the Uniroyal ball, Wilson's claim 1 could not be stretched to cover Dunlop's balls.

Why the Dependent Claims Still Mattered Normally, if the main claim is not infringed, the narrower claims built on it are not infringed either. Here, though, claim 1 failed only because of the prior art, so the narrower claims needed their own review because their extra details might avoid the old ball.

The Dependent Claims Also Fail The court reviewed each of the narrower claims, each of which is a small variation on the six-circle ball. It concluded that none could reach Dunlop's balls without going beyond what Wilson could lawfully have patented.

The Court's Conclusion The trial judge was wrong to deny Dunlop's request to overturn the verdict, because any range broad enough to cover Dunlop's balls would also cover the prior art. The appeals court therefore reversed the finding that Dunlop infringed.

Indirect Infringement

Why Indirect Infringement Exists The Supreme Court explained that this doctrine stops people from undermining patents by helping others infringe instead of infringing themselves. This protection is especially important when it is hard to go after the people who actually do the infringing. It lets patent owners go after people who help someone else infringe, such as by supplying a part or giving instructions.

Lucent Technologies, Inc. v. Gateway, Inc. 580 F.3d 1301 (Fed. Cir. 2009) Background

What the Day Patent Covers AT&T engineers applied for the patent in 1986, and it covers a way to fill in fields on a computer screen without using a keyboard. Users fill in fields with on-screen tools such as a graphical keyboard, a menu, a calculator, or a stylus.

Why Microsoft Could Only Be an Indirect Infringer The claims cover a method, so simply selling software cannot infringe them. Infringement happens only when a user performs the method, so Microsoft could be liable only for contributing to or inducing that infringement.

Procedrual Posture

The Lucent Trial Lucent claimed that Microsoft Money, Outlook, and Windows Mobile infringed the patent. The jury found Microsoft liable, and Microsoft appealed after its post-trial challenges failed.

The Lucent Appeal Microsoft appealed a jury verdict finding that it indirectly infringed Lucent's "Day patent". The appeals court upheld the verdict because it followed the law and was supported by enough evidence.

Legal Standard

The Test for Contributory Infringement A seller contributes to infringement if it sells a part made especially for an infringing use, knowing this, and the part has no substantial use that doesn't infringe. The patent owner must also prove that someone actually infringed and that the seller knew the combination was patented and infringing.

The Key Dispute: What Counts as the "Product"? Microsoft argued its software as a whole had many uses that don't infringe, while Lucent argued that the date-picker tool had none. If the product is the whole software package, Microsoft wins, but if it is the date-picker alone, Lucent wins.

Legal Reasoning

The Five-Feature Example The court imagined a program whose five features each infringed a different patent. Under Microsoft's view, the seller could never be liable because the program as a whole has uses that don't infringe, which the court found unworkable and inconsistent with the law.

You Can't Hide an Infringing Part Inside a Bigger Product If Microsoft had sold the date-picker as a separate download, it would clearly have been contributing to infringement. Under an earlier case, a seller cannot avoid liability just by building the infringing part into a larger product.

Microsoft's Reliance on Hodosh Microsoft read the Hodosh case to mean courts should look at the whole product sold, not one ingredient. The court had already rejected that reading, because Hodosh does not let otherwise infringing parts escape liability just by being combined with other parts.

What Hodosh Actually Held Hodosh involved a toothpaste containing potassium nitrate, used in a patented method for desensitizing teeth. Potassium nitrate has many uses when sold by itself, but inside that toothpaste it was suitable only for the infringing use.

Applying Hodosh to the Date-Picker Likewise, the date-picker is suitable only for infringing, and putting it inside a larger program does not change that. The jury could reasonably find that Microsoft meant for users to use the tool, and the tool's only intended use infringed.

Global-Tech Appliances, Inc. v. SEB S.A. 563 U.S. 754 (2011) Background

The Global-Tech Question The Supreme Court considered whether a party that "actively induces" infringement must know that the acts it encourages infringe a patent. The case involved SEB, a French company that invented a "cool-touch" deep fryer.

SEB's Fryer SEB's fryer kept its outer surface cool by separating the hot pot from the plastic housing with an insulating ring. SEB patented the design in the United States in 1991, and it sold very well under the T-Fal brand

Pentalpha Copies the Fryer Pentalpha, a Hong Kong company, bought an SEB fryer made for overseas sale, which had no U.S. patent markings, and copied everything but its looks for Sunbeam. It then hired a lawyer to check for patent problems but did not tell him the design was copied.

Sales and Lawsuits The lawyer did not find SEB's patent, and Pentalpha sold its fryers to Sunbeam, which undercut SEB on price in the U.S. market. Even after Sunbeam told Pentalpha it had been sued by SEB, Pentalpha kept selling fryers to other U.S. resellers.

Procedural Posture

SEB Sues Pentalpha SEB sued Pentalpha both for infringing directly and for inducing its customers to infringe. The jury ruled for SEB on both claims and found that the infringement was willful

Pentalpha's Defense and the Lower Courts' Ruling Pentalpha argued that it could not have induced infringement because it did not know about the patent until April 1998. The lower courts disagreed, finding that Pentalpha deliberately ignored a known risk that a patent existed and treating that as a form of knowledge.

Legal Standard

The Statute's Wording Is Unclear The statute simply says that anyone who "actively induces infringement" is liable, and it does not mention intent. Even so, the Court concluded that at least some intent is required, since "actively induce" implies taking deliberate steps toward a result.

The Used Car Example Saying a salesman "induced a customer to buy a damaged car" could mean he knew about the damage or just that the car happened to be damaged. The statute is unclear in the same way, so it could require only encouraging conduct that happens to infringe, or knowing that the conduct infringes.

Turning to Older Cases Because the wording settled nothing, the Court looked at cases from before the 1952 Patent Act, which the statute was meant to write into law. Those cases were unclear, and they treated both inducing and supplying parts as one broad idea of "contributory infringement".

The Aro II Precedent The Court found its earlier Aro II decision decisive, because Aro II required a parts supplier to know the combination was patented and infringing. The contributory infringement provision has the same unclear wording, and a closely divided Court in Aro II still required knowledge of the patent.

Court Reasoning

Applying the Same Rule to Inducement Aro II has become settled law, SEB did not ask the Court to overturn it, and Congress has not changed it. Because the two provisions share the same origin and the same wording problem, the Court held that inducement also requires knowing that the encouraged acts infringe a patent.

The Wrong Standard, but the Right Result The Court rejected the lower court's "deliberate indifference to a known risk" standard. It still upheld the verdict, because the evidence showed Pentalpha's knowledge under the doctrine of "willful blindness".

What Willful Blindness Is Criminal law has long held that people cannot avoid liability by deliberately avoiding clear evidence of key facts, because they are just as guilty as people who actually know. The Court saw no reason not to apply this idea to induced patent infringement, especially since Pentalpha clearly knew its customers were selling the fryers in the United States.

The Two Requirements for Willful Blindness A person is willfully blind if they believe a fact is very likely true and take deliberate steps to avoid learning it. This is a higher bar than recklessness, which is simply knowing about a risk, or carelessness, which is failing to notice a risk one should have seen.

Where the Lower Court Went Wrong The lower court's test was too loose, because it accepted a mere "known risk" and did not require active efforts to avoid knowing. Under the correct test, however, the evidence still showed that Pentalpha willfully blinded itself to the infringement.

The Evidence Against Pentalpha Pentalpha copied an innovative, best-selling fryer and chose an overseas model, knowing such models usually lack U.S. patent markings. Its CEO also hid the copying from the patent lawyer, apparently to set up a "plausible deniability" defense, and this was more than enough evidence for the jury.

Akamai Technologies, Inc. v. Limelight Networks, Inc. 797 F.3d 1020 Background

Akamai Returns From the Supreme Court The Supreme Court sent the Akamai case back, suggesting the appeals court may have read the direct infringement rule too narrowly. The full appeals court unanimously restated the law and found enough evidence that Limelight directly infringed Akamai's patent, reversing the trial court's ruling of no infringement.

Legal Standard

When Multiple Parties Perform the Steps Direct infringement of a method patent happens only when one party performs, or is responsible for, every step. When several parties are involved, one party is responsible for the others' steps if it directs or controls them or if they form a joint enterprise.

What "Direct or Control" Means A party is responsible for steps performed by its agents or by people it hires by contract to perform them. A party is also responsible when it makes performing the steps a condition of using its service or getting a benefit and sets how or when the steps are done.

Who Decides Direction or Control When that happens, the other parties' actions are treated as the controlling party's own, making it the single infringer. Whether a party directed or controlled others is a factual question for the jury, and on appeal the court asks only whether there was enough evidence.

The Joint Enterprise Route When parties form a joint enterprise, each is treated as responsible for the others' steps, as if they were one actor. A joint enterprise requires an agreement, a shared purpose, a shared financial interest, and an equal say in how the enterprise is run.

Why the Court Chose This Approach The court said this approach fits best with the statute's wording, its purpose, and earlier cases. It rejected the earlier panel's view that only agency, contract, and joint enterprise relationships count, holding that the real question is whether all the steps can be credited to one party.

Room for Future Cases The court said other situations may come up in the future where one party should be held responsible for others' steps. Going forward, courts will decide these questions based on the specific facts of each case

Procedural Posture

The Akamai Lawsuit In 2006, Akamai sued Limelight for infringing a patent on methods for delivering content over the internet. Both sides agreed that Limelight's customers, not Limelight itself, performed the "tagging" and "serving" steps.

The Jury Verdict and Its Reversal The jury was told Limelight was responsible if it directed or controlled its customers, and it found that Limelight infringed. The trial court at first upheld the verdict, but after the Muniauction decision it reversed itself and ruled that Limelight could not be liable as a matter of law.

Legal Reasoning

The Court Reinstates the Verdict The appeals court reversed and restored the jury's verdict, finding enough evidence that Limelight directed or controlled its customers' steps. Specifically, Limelight made its customers perform those steps to use its network and controlled how or when they did them.

The Contract Requires Serving The contract says Limelight is not responsible if its service fails because a customer does not serve the content. So customers who want to use Limelight's service must tag and serve content, which shows Limelight made those steps a condition of using the service.

Limelight Controls How Steps Are Done After a deal is signed, Limelight sends a welcome letter explaining how to use its service and naming a Limelight manager to lead the setup. Limelight also provides a hostname and step-by-step instructions, and the service won't work unless customers follow those steps exactly.

Ongoing Involvement by Limelight's Engineers Limelight's engineers help customers set up the service, test its quality, and stay available to fix problems. Customers therefore do not act on their own, because Limelight controls how and when they perform the steps.

Conclusion Because all of the steps could be credited to Limelight, the court held Limelight liable for direct infringement and reversed the trial court's ruling. It sent the case back to a smaller panel of judges to resolve the remaining issues.