The Ethical & Practical Use of AI in Patent Prosecution

The Ethical and Practical Use of AI in Patent Prosecution

Washington, D.C., September 4, 2025 — As artificial intelligence (AI) reshapes the legal landscape, its integration into patent prosecution is transforming how attorneys draft applications, conduct prior art searches, and respond to office actions. While AI offers significant efficiencies, its use raises ethical and practical challenges that demand careful navigation to ensure fairness, transparency, and compliance with legal standards. At the recent Law.com Intellectual Property Summit, experts explored how patent practitioners can responsibly leverage AI to enhance productivity while addressing risks like bias, accountability, and regulatory compliance.

Practical Benefits of AI in Patent Prosecution

AI is streamlining labor-intensive tasks in patent prosecution, enabling attorneys to focus on strategic, high-value work. Key applications include:

• Enhanced Prior Art Searches: AI-powered tools using natural language processing (NLP) and machine learning, such as those from PowerPatent, analyze vast patent databases and non-patent literature faster and more comprehensively than traditional methods. These tools can identify relevant prior art with greater precision, reducing the risk of overlooking critical references.
• Automated Drafting: Large language models (LLMs) like those developed by Solve Intelligence can generate initial drafts of patent specifications, claims, and responses to office actions in seconds, cutting preparation time by up to 50%. This efficiency lowers costs for clients, particularly individual inventors and small entities.
• Office Action Analysis: AI systems suggest amendments or arguments based on successful responses in similar cases, helping attorneys craft effective strategies to overcome rejections. They also identify patterns in examiner behavior, enabling tailored responses that align with specific art units.
• Portfolio Management: AI assists in analyzing market trends, assessing patent value, and identifying strategic opportunities, allowing firms to optimize clients’ intellectual property portfolios.

These tools have driven measurable gains. A 2023 McKinsey report highlighted the legal sector’s potential for productivity increases through AI adoption, with patent practices seeing reduced drafting times and improved accuracy. The U.S. Patent and Trademark Office (USPTO) itself has embraced AI, incorporating tools like the Patents End-to-End (PE2E) search for prior art analysis, enhancing examiner efficiency.

Ethical Considerations and Challenges

While AI’s benefits are substantial, its use in patent prosecution raises significant ethical concerns that practitioners must address to maintain the integrity of the patent system:

• Bias and Data Quality: AI systems are only as reliable as their training data. In patent law, where many documents remain confidential for months or years, keeping AI tools updated with current laws, regulations, and case law is challenging. Outdated or biased data can lead to flawed applications or missed prior art, risking rejections or invalid patents.
• Accountability and Liability: The impersonal nature of AI complicates accountability. If an AI-generated application contains errors or omissions leading to rejection or an unenforceable patent, determining responsibility—whether it lies with the attorney, the AI provider, or the client—remains unclear. This ambiguity can leave clients vulnerable without clear recourse.
• Transparency and Duty of Candor: The USPTO’s April 2024 guidance emphasizes that practitioners must uphold their duty of candor and good faith under 37 CFR 1.56(a), disclosing all material information related to patentability. Attorneys using AI must fact-check outputs for accuracy, as LLMs like ChatGPT are known to occasionally produce errors or “hallucinations.” Failure to verify AI-generated content risks violating ethical obligations and jeopardizing filings.
• Client Confidentiality: AI tools often process sensitive client data, raising concerns about data security and compliance with privacy regulations. The USPTO’s guidance stresses the need for robust security architectures, encryption, and compliance certifications when selecting AI vendors to protect client information.

Regulatory Guidance and Best Practices

The USPTO has taken proactive steps to address AI’s role in patent prosecution. Its April 2024 notice clarifies that practitioners can use generative AI within the agency’s code of conduct, provided they adhere to existing rules. Key requirements include:

• Signature Certification: Under 37 CFR 11.18(b)(1), attorneys must sign all USPTO submissions, certifying their accuracy after reasonable inquiry. This necessitates thorough review of AI-generated documents to avoid misstatements.
• Inventorship Standards: The USPTO’s February 2024 guidance on AI-assisted inventions reaffirms that only human inventors can be named, per Thaler v. Vidal (2022), but humans using AI can qualify as inventors if they make significant contributions, as defined by the Pannu factors. These factors require contributions to the conception or reduction to practice of the invention that are not insignificant in quality.

To navigate these challenges, experts recommend:

• Human Oversight: Attorneys must critically review AI outputs to ensure accuracy and compliance with USPTO standards. Automated drafts should be treated as starting points, not final submissions.
• Transparent Use: While there’s no obligation to disclose AI use, practitioners should document their processes to demonstrate due diligence if questioned.
• Vendor Vetting: Select AI tools with robust security measures and transparent algorithms to mitigate risks of data breaches or unreliable outputs.
• Continuous Training: Firms should invest in training to help attorneys understand AI’s capabilities and limitations, fostering responsible adoption.

Looking Ahead

AI’s integration into patent prosecution offers transformative potential, from cost savings to enhanced efficiency, but its ethical and practical challenges require careful management. As the USPTO continues to refine its guidance and courts address AI-related disputes, practitioners must stay informed to balance innovation with accountability. By adopting robust oversight and ethical frameworks, patent attorneys can harness AI to deliver better outcomes for clients while upholding the integrity of the patent system.

For more information, contact the USPTO Office of Public Affairs at opa@uspto.gov or visit www.uspto.gov.