This third episode of the PatentRiff Podcast is here and we’re talking about AI and patent practice.
The podcast discusses:
AI in Patent Practice Risks Patent Bars: Using AI tools for invention details could potentially inadvertently trigger public use or disclosure bars (35 U.S.C. § 102) if confidentiality isn't guaranteed, especially with public-facing AI.
Ethical Obligations: Attorneys must uphold confidentiality, privilege, and the duty of candor when using AI.
USPTO Guidance (2024): The USPTO emphasizes natural persons signing submissions, certifying accuracy, and reviewing AI-generated content, noting AI cannot be an inventor.
Practitioners Bear Responsibility for AI Use: The USPTO's Office of Enrollment and Discipline (OED) can sanction practitioners for rule violations, including submitting inaccurate or fabricated information via AI, akin to supervising a non-practitioner assistant.
Practical Challenges with Legal AI Tools: Current AI tools present issues like the "black box" problem (lack of transparency), prompt engineering difficulties, the need for multiple specialized tools rather than a single comprehensive one, and limitations in scaling for IP portfolio analysis.
Cautious Adoption and Human Oversight are Paramount: Despite the promise of AI, practitioners must exercise extreme caution, vet AI tools diligently, understand data handling, and maintain rigorous human oversight and review to navigate evolving legal and technological landscapes.
Share this post