Quantum Computing Patent Enforcement: The Coming Challenges

Need to know

The last decade of quantum computing patenting has produced a large and steadily growing stock of issued U.S. patents, yet quantum computing patent infringement suits are rare. As the field matures and infringement cases increase, they may present unique challenges, including the difficulty of reverse engineering quantum computing technologies, the susceptibility of quantum computing patents to patent eligibility challenges, and the problem of proving infringement of decentralized, cloud-based quantum computing technologies.  

---

Quantum computing has been among the most active patent races across all technologies. Large technology companies, startups, and research institutions have built substantial patent portfolios, and U.S. quantum computing filings grew roughly ten times from 2016 to 2023. Yet quantum computing patent enforcement remains largely untested. The few suits that have emerged sit at the field’s edges, and no case appears to have tested infringement of core quantum operations on the merits.

Nevertheless, when infringement cases do reach the core technology, they are likely to present unique challenges. On one hand, infringement of foundational patents — those covering the creation, control, and manipulation of qubits themselves — cannot easily be proved through reverse engineering, as their operations are based on quantum states that generally cannot be observed from outside. On the other, patents drafted at a more detectable level may face invalidity challenges, as mere applications of quantum computing or characterizations of the quantum computers as black boxes may invite abstract-idea scrutiny under § 101. 

The strongest patents on paper may thus prove the hardest to enforce, while the most enforceable may prove the most vulnerable to invalidation. Thus, the most valuable patents in practice are likely to lie somewhere between those two extremes. 

The fundamentals of quantum computing 

A classical bit is always 0 or 1. In quantum computing, the basic unit of information is called a qubit. Loosely speaking, a qubit is a probabilistic combination of 0 and 1 at the same time, a condition known as superposition. At the heart of quantum technologies is entanglement, which links two or more qubits so that they behave as an interdependent system rather than as independent particles. Measuring one qubit reveals outcomes correlated with those of the others. Entangled qubits can represent many combinations of values at once, which is useful for certain kinds of computation. As a result, quantum computers are expected to solve certain problems, like breaking widely used data encryption schemes or simulating molecular behavior, far faster than any classical machine. 

Quantum computing hardware uses diverse physical systems to create and manipulate qubits. Mainstream and actively pursued approaches are based on a variety of technologies, including superconducting circuits, trapped ions, neutral atoms, photonics, and spins of atomic nuclei. Each method has its own advantages and tradeoffs. For enforcement purposes, however, certain physical consequences of quantum mechanics matter more than any hardware distinction.

One example is measurement collapse. Before measurement, a qubit may exist in a superposition represented mathematically by a wave function representing probabilities associated with both 0 and 1. Ordinary readout of a qubit is what physicists call “projective measurement.” This measurement forces the qubit into a definite 0 or 1. The readout collapses the system’s wave function — its full mathematical description — into a single classical result.  

Quantum mechanics forbid making a perfect copy of an arbitrary unknown quantum state, such as the live state of a qubit in mid-computation under the so-called “no-cloning theorem.” No known instrument, consistent with standard quantum mechanics, can perfectly copy an arbitrary unknown quantum state. 

The reverse engineering challenge  

Traditional patent enforcement rests on a practical premise: The patentee can obtain the accused product and take it apart. Electrical circuits can be traced. Chipsets can be delayered and imaged. Software can be tested or even decompiled. Teardown of consumer products is a staple process in pre-suit infringement investigations.

Quantum computing complicates that premise in two ways:

• First, there is no mass-market product to purchase and tear down. Quantum hardware is expensive, complex, and proprietary. It may be built around cryogenic refrigeration, precision laser systems, and/or custom control electronics rather than off-the-shelf servers. The machines typically sit in provider facilities and are reached remotely.
• Second, physics limits what inspection could reveal even with access. Consider claims directed to internal quantum operations, such as an entanglement scheme or an error-correction method acting on live qubits. Measurement collapse means the operation generally cannot be watched while it happens or reconstructed afterward. In sum, direct observation is often unavailable as a matter of physics. The result can be a patent that is valid yet difficult or impractical to enforce.  

The patent eligibility challenge

The observable spectrum of patent claims in quantum computing would cover, on one end, highly nuanced physical implementations of quantum technologies, which are hard to enforce, and, on the other end, applications of quantum computing that treat the quantum computer as a black box and claim only its inputs and outputs. Patents of the second type may be easier to enforce, but they could be more vulnerable to invalidity attacks, including eligibility challenges under § 101, as being directed to unpatentable abstract ideas.

For patent claim drafting, the target therefore lies between these two extremes. High-quality, enforceable patents are directed to practical implementations of quantum technology and claimed at a level in which a reasonable investigation can establish use.

For innovations that cannot be reverse engineered, and for approaches that may not survive patent validity scrutiny, trade secret protection may be the better option. The same opacity that frustrates patent enforcement can make trade secrets unusually durable. Trade secrets, however, carry their own risks, like misappropriation or lack of protection against independent development. They can also complicate licensing and financing. Patents, by contrast, retain deterrent and cross-licensing value even when direct enforcement is uncertain. 

The divided infringement challenge

Another unique aspect of quantum computing technologies is that access to them is likely to be cloud- or network-based. Essentially, quantum computing is or will be offered as software as a service. Two enforcement consequences follow. First, there is no physical product a patentee can buy and inspect before filing suit. And second, the accused system may be operated by one party, accessed by another, and located anywhere the provider keeps its machines, including outside the United States.

That landscape makes the choice between method and system claims unusually consequential. A method claim is directly infringed only when every step is performed within the U.S. That requirement is fragile when the executing hardware may sit abroad. Method claims also face a divided-performance problem. Steps performed by multiple actors are attributed to a single actor only in narrow scenarios, e.g., when one actor directs or controls the others' performance, or in case of a joint enterprise. Quantum computing as a service could therefore present a divided infringement problem.  

Indirect infringement theories offer fallback positions, but they have limitations. Inducement requires knowledge of the patent and intent to encourage infringement. Contributory infringement requires a component with no substantial non-infringing use. Both induced and contributory infringement require an underlying act of direct infringement. That requirement circles back to the single actor problem above.

System claims can fare better. A domestic customer who puts a claimed system into service — such as by controlling the system as a whole and obtaining a benefit from it — may be a viable single direct infringer even when back-end hardware is located elsewhere. 

Takeaways

Quantum computing patent enforcement challenges are manageable but require attention long before disputes arise.

In prosecution:

• Draft claims around what can be detected and proved, such as inputs, outputs, control, calibration, benchmarks, and interfaces, rather than internal quantum states. A limitation no one can observe is one that is difficult to prove.
• Draft single-actor claims where possible. For cloud-based implementations, consider claims directed separately to the user, the cloud platform, or the hardware operator.
• Weigh patent and trade secret protection deliberately for innovations that cannot be reverse engineered and are unlikely to be detectable through ordinary investigation.

In litigation:

• In pre-suit diligence, map the geography and the actors. Confirm the U.S. nexus and identify a single direct infringer, such as a domestic customer who uses the complete claimed system.
• Use system-claim theories where available. A domestic customer who controls and benefits from the claimed system may supply the single infringer that method claims lack.
• Consider building the infringement case on evidence outside the physics of quantum processing.  
• On defense, implement controls for records and publications.