UPDATE – FEBRUARY 2026:
The U.S. Patent and Trademark Office’s AI approach has continued to evolve with several policy and examination shifts reflecting broader changes in federal AI priorities. While the April 2024 practitioner guidance remains in effect—especially regarding duties of candor, confidentiality, reasonable inquiry, and responsible use of AI tools—the agency has adjusted its broader policy direction under new leadership and executive guidance.
In late 2025, the USPTO issued updated AI inventorship guidance reaffirming that only natural persons may be named as inventors, clarifying that AI systems are treated as tools rather than legal inventors. This update aligned USPTO policy with evolving federal executive orders emphasizing innovation and reduced regulatory barriers for AI development. The agency also revised examination guidance related to patent eligibility (§101), signaling a more flexible approach toward AI-related inventions and cautioning examiners against automatically categorizing AI claims as abstract mental processes.
Additional internal memoranda released in late 2025 refined patent examination practices to support clearer, more consistent treatment of AI-enabled technologies. These changes reflect a broader shift toward encouraging AI innovation while maintaining existing legal safeguards and disclosure obligations. The USPTO’s January 2025 Artificial Intelligence Strategy remains an important framework, but the office has indicated that the strategy is under ongoing review to ensure alignment with updated federal priorities and agency leadership direction.
As of early 2026, no new overarching AI guidance has replaced the April 2024 memo. Instead, the USPTO continues to build on that foundation through targeted updates to inventorship policy, patent eligibility interpretation, and examiner training. Practitioners should continue to treat the 2024 guidance as the baseline for responsible AI use in USPTO proceedings while monitoring evolving policy adjustments that increasingly favor innovation-friendly treatment of AI technologies.
ORIGINAL NEWS STORY:
U.S. Patent and Trademark Office Issues Guidance on AI
The United States Patent and Trademark Office (USPTO) recently issued guidance on AI. The document, released on April 10, informs practitioners and the public about important issues to navigate when using AI in matters before the USPTO. While the USPTO is committed to maximizing AI’s benefits, it also recognizes risks that need mitigation through technical solutions and human governance.
Adapting to New Rules
The USPTO has determined that existing rules protect its ecosystem against potential AI perils. This guidance reminds individuals involved in USPTO proceedings of pertinent rules and policies, informs them of risks associated with using AI systems, and provides suggestions to mitigate those risks. A key existing rule is the duty of candor and good faith. Individuals associated with a USPTO proceeding have a duty of candor, good faith, and disclosure of material information. This duty extends to actions taken with AI tools. The duty is broader than just the duty of disclosure – it also prohibits acts of fraud, intentional misconduct, and making incorrect statements to the USPTO.
Another important rule relates to signature requirements. Nearly all correspondence filed at the USPTO must bear a natural person’s signature. By providing this signature, the person is certifying that the statements in the document are true and that they performed a reasonable inquiry into the contents. Simply relying on AI output, without reviewing and verifying it, does not satisfy this reasonable inquiry requirement.
Practitioners must also follow confidentiality obligations under the USPTO’s rules. Specifically, practitioners must maintain client confidentiality and take reasonable steps to prevent unauthorized disclosure of client information, including inadvertent disclosure to third-party AI systems. Use of AI tools risks improperly divulging sensitive client information.
Foreign Filing and Export Regulations
When it comes to foreign filing licenses and export regulations, practitioners must be mindful that using AI tools abroad or tools that transmit data internationally may implicate foreign filing license requirements under U.S. laws as well as export control regulations.
Additionally, the USPTO has policies governing authorized access and use of its electronic filing systems and websites. Users must follow these policies, which state that only authorized individuals can file documents or access certain information. AI tools themselves are not considered “users” that can obtain USPTO accounts or credentials under these policies.
The guidance further explains how these key existing rules and policies apply in specific contexts when using AI tools before the USPTO, such as drafting documents, preparing filings, accessing IT systems, and upholding duties to clients. The guidance also highlights confidentiality risks of disclosing information to third-party AI systems located abroad, potential deemed exports, and national security implications.
Conclusion
Fraud or intentional misconduct involving AI will not be tolerated and violates the duty of candor. Unauthorized access or modification of USPTO data via AI constitutes computer fraud. The USPTO aims to protect against AI-related harms while promoting responsible AI use aligned with its rules and policies. It will continue studying AI’s impacts through initiatives like the Artificial Intelligence and Emerging Technologies Partnership. Practitioners should carefully review existing obligations before using AI tools in USPTO matters.
Need Help?
Keeping track of the everchanging AI landscape can be tough, especially if you have questions and concerns about how it will impact you. Don’t hesitate to reach out to BABL AI. Their Audit Experts are ready to provide valuable assistance.
