[Updated USPTO Guidance]July 2024: Understanding patent eligibility can be a challenging task, especially with the ever-evolving landscape of technology. The United States Patent and Trademark Office (USPTO) has released new Subject Matter Eligibility Examples in July 2024, focusing on the application of artificial intelligence (AI) and other advanced technologies (available online here). These examples aim to clarify how the USPTO evaluates patent claims under Section 101 of the U.S. Patent Act, particularly in the context of AI-driven innovations.
Summary of Guidance Update and Impact on Examination Procedure and Prior Examination Guidance
Sections of the guidance update provide an overview of the USPTO’s existing patent subject matter eligibility guidance and specific updates for AI inventions. These sections cover:
- Evaluation of Abstract Ideas: Determining if a claim recites an abstract idea (Step 2A, Prong One).
- Practical Application: Assessing if the claim integrates a judicial exception into a practical application, improving the functioning of a computer or another technology (Step 2A, Prong Two).
- AI-Assisted Inventions: Addressing the eligibility of AI-assisted inventions.
- Examples 47-49: Introducing examples to help examiners apply the guidance to AI inventions during the patent examination process.
The update reinforces that while it focuses on AI inventions, portions of the guidance apply to other types of inventions. The guidance aims to ensure consistency with existing USPTO practices and is intended to be incorporated into the Manual of Patent Examining Procedure (MPEP).
Overview of the USPTO’s Patent Subject Matter Eligibility Guidance
The USPTO’s subject matter eligibility guidance is found in MPEP sections 2103-2106.07(c) and is used to analyze claims across all technologies. The guidance combines the criteria for eligibility into a single analysis for all categories of claims and types of judicial exceptions.
Step 1: Determine if the claimed invention falls into at least one of the four categories recited in 35 U.S.C. 101. If no, the claim is ineligible. If yes, the claim is evaluated under Step 2.
Step 2: Apply the Supreme Court’s two-part framework (Alice/Mayo Steps 1 and 2) to identify claims directed to a judicial exception and evaluate if additional elements provide an inventive concept. If the answer to Step 2 is no, the claim is eligible. If yes, assess if the claim integrates the exception into a practical application. If it does, the claim is eligible; if not, it is ineligible.
USPTO Provided Examples 47 and 48
Example 47: Anomaly Detection Using AI
Claim 1: Application Specific Integrated Circuit (ASIC) for an Artificial Neural Network (ANN)
- Claim Language:
An application-specific integrated circuit (ASIC) for an artificial neural network (ANN, a type of AI model), the ASIC comprising:
a plurality of neurons organized in an array, wherein each neuron comprises a register, a microprocessor, and at least one input; and
a plurality of synaptic circuits, each synaptic circuit including a memory for storing a synaptic weight, wherein each neuron is connected to at least one other neuron via one of the plurality of synaptic circuits.
- Explanation: This claim describes a physical ASIC for an ANN, detailing the hardware components like neurons, registers, microprocessors, and synaptic circuits.
- Eligibility: The claim is patent eligible because it falls within the statutory category of a machine. It does not recite any judicial exceptions (abstract ideas, laws of nature, or natural phenomena), focusing instead on a specific hardware implementation.
Claim 2: Method of Using an Artificial Neural Network (ANN)
- Claim Language:
A method of using an artificial neural network (ANN) comprising:
(a) receiving, at a computer, continuous training data;
(b) discretizing, by the computer, the continuous training data to generate input data;
(c) training, by the computer, the ANN based on the input data and a selected training algorithm to generate a trained ANN, wherein the selected training algorithm includes a backpropagation algorithm and a gradient descent algorithm;
(d) detecting one or more anomalies in a data set using the trained ANN;
(e) analyzing the one or more detected anomalies using the trained ANN to generate anomaly data; and
(f) outputting the anomaly data from the trained ANN.”
- Explanation: This claim involves steps of receiving and discretizing continuous training data, training an ANN, detecting and analyzing anomalies, and outputting anomaly data.
- Eligibility: The claim is not patent eligible. It recites judicial exceptions, including mathematical concepts and mental processes (steps performed by the computer can be performed mentally), and does not integrate these exceptions into a practical application. Additionally, it does not provide an inventive concept beyond the abstract idea of using ANN to receive and analyze data.
Claim 3: Method of Using an ANN to Detect Malicious Network Packets
- Claim Language:
A method of using an artificial neural network (ANN) to detect malicious network packets comprising:
(a) training, by a computer, the ANN based on input data and a selected training algorithm to generate a trained ANN, wherein the selected training algorithm includes a backpropagation algorithm and a gradient descent algorithm;
(b) detecting one or more anomalies in network traffic using the trained ANN;
(c) determining at least one detected anomaly is associated with one or more malicious network packets;
(d) detecting a source address associated with the one or more malicious network packets in real time;
(e) dropping the one or more malicious network packets in real time; and
(f) blocking future traffic from the source address.”
- Explanation: This claim involves training an ANN, detecting anomalies in network traffic, determining if anomalies are malicious, detecting source addresses, dropping malicious packets, and blocking future traffic.
- Eligibility: The claim is patent eligible because it integrates the judicial exception (the abstract idea of anomaly detection) into a practical application by improving network security. This method provides specific improvements in the field of network intrusion detection and remediation, thus demonstrating a practical application of the abstract idea.
Example 48: Speech Separation Technology
Claim 1: Speech Separation Method
- Claim Language:
A speech separation method comprising:
(a) receiving a mixed speech signal x comprising speech from multiple different sources sn, where n ∈ {1, . . . N};
(b) converting the mixed speech signal x into a spectrogram in a time-frequency domain using a short time Fourier transform and obtaining feature representation X, wherein X corresponds to the spectrogram of the mixed speech signal x and temporal features extracted from the mixed speech signal x; and
(c) using a deep neural network (DNN, a type of ANN) to determine embedding vectors V using the formula V = fθ(X), where fθ(X) is a global function of the mixed speech signal x.”
- Explanation: This claim describes receiving a mixed speech signal, converting it into a spectrogram, and using a DNN to determine embedding vectors based on the signal.
- Eligibility: The claim is not patent eligible. It recites mathematical concepts and abstract ideas without integrating them into a practical application. It does not provide an inventive concept beyond the abstract idea of using mathematical operations for speech separation.
Claim 2: Enhanced Speech Separation Method
- Claim Language:
The speech separation method of claim 1 further comprising:
(d) partitioning the embedding vectors V into clusters corresponding to the different sources sn;
(e) applying binary masks to the clusters to create masked clusters;
(f) synthesizing speech waveforms from the masked clusters, wherein each speech waveform corresponds to a different source sn;
(g) excluding speech signals from unwanted sources; and
(h) transmitting the synthesized speech waveforms.”
- Explanation: This claim builds on Claim 1 by adding steps to partition embedding vectors into clusters, apply binary masks, synthesize speech waveforms, exclude certain speech signals, and transmit the result.
- Eligibility: The claim is patent eligible because it integrates the abstract idea into a practical application. It demonstrates an improvement in speech separation technology by addressing specific problems in separating speech from multiple sources and providing a solution that enhances the accuracy and quality of the separated speech signals.
Claim 3: Computer-Readable Storage Medium for Speech Separation
- Claim Language:
A non-transitory computer-readable storage medium having instructions stored thereon that, when executed by a processor, cause the processor to perform the method of claim 2.”
- Explanation: This claim involves a non-transitory computer-readable storage medium with instructions to perform the speech separation method.
- Eligibility: The claim is patent eligible because it integrates the abstract idea into a practical application by improving speech-to-text transcription. The method provides a concrete solution to the problem of accurately transcribing speech in noisy environments, demonstrating a practical application of the underlying abstract idea.
Conclusion
The USPTO’s examples emphasize the importance of demonstrating how an invention applies abstract ideas in a practical and innovative way. Claims that merely involve abstract concepts or mathematical calculations without practical application or inventive concepts are likely to be deemed ineligible. However, when these abstract ideas are integrated into practical applications that improve technology or offer tangible benefits, they can become eligible for patent protection. These detailed examples serve as a valuable guide for inventors and patent practitioners, illustrating the nuances of subject matter eligibility in advanced technological fields.
Michael Jones is the managing member at Jones Intellectual Property. His practice specializes in all aspects of intellectual property, including patent, trademark, and copyright law. He can be reached at mjones@jonesipl.com.