The European Patent Institute (epi) has just introduced new guidelines on the use of generative AI for all European patent attorneys. As artificial intelligence continues to reshape our profession, these guidelines serve as a critical framework to ensure we continue prioritizing professional integrity, confidentiality, and the interests of our clients.
This article breaks down key takeaways from the guidelines and provides a concise Q&A to address common concerns.
Overarching Principle: Client Interests First
The epi emphasizes:
“When using AI of any kind in professional work, a Member must adopt the highest possible standards of probity; must take all reasonable steps to maintain confidentiality when this is required; and at all times must put the interests of clients first as required by Article 1 of the epi Code of Conduct.”
epi Guidelines, “Overarching Principles”
This principle underpins every aspect of the guidelines. Regardless of the tools we use, the cornerstone of our profession remains the responsibility to deliver accurate, high-quality work while maintaining the trust and interests of our clients.
Breaking Down the Guidelines
1. Understanding Generative AI
Patent attorneys are urged to thoroughly understand the capabilities and limitations of generative AI models. Key risks include “hallucinations” (false outputs) and confidentiality breaches. Continuous learning about these technologies is essential for responsible use.
Did you know that my course, “Confidential Patent Drafting with Generative AI: A Beginner’s Guide for Patent Professionals,” offers a practical introduction to understanding AI tools, addressing their risks, and integrating them confidently into your practice?
2. Client Transparency
Patent attorneys must establish, in advance of using generative AI, the wishes of their clients with regard to the use of generative AI. The guidelines recommend to keep records of enquiries sent to clients in order to establish their preferences with regard to generative AI, and the responses to those enquiries.
3. Confidentiality and Data Security
Protecting client information is paramount. Attorneys should verify the confidentiality policies of AI models and use secure systems to handle sensitive data. If a model’s security is unclear, it should not be used for confidential tasks.
4. Accuracy and Responsibility
Patent attorneys remain fully accountable for their work, irrespective of AI involvement. Outputs generated by AI must be meticulously reviewed and meet the same professional standards as work produced without AI tools. Patent attorneys cannot cite the use of generative AI as any excuse for errors or omissions.
Questions and Answers
1. What is generative AI, and how does it differ from non-generative AI?
Generative AI creates new content, such as text or images, based on training data. Non-generative AI, by contrast, analyzes existing data to identify patterns or make recommendations (e.g., spell checkers or translation tools).
2. Why is understanding generative AI crucial for patent attorneys?
Generative AI poses unique risks like hallucinations and confidentiality issues. Attorneys need to be aware of these limitations to prevent errors and maintain the highest quality standards. As the quality of AI output heavily depends on the quality of your inputs, my Prompt Engineering Masterclass teaches you how to craft precise, effective prompts tailored for patent drafting, ensuring better results and minimizing risks.
3. Are patent attorneys required to disclose the use of AI to authorities?
Guideline 5b states that patent attorneys are not currently required to disclose the use of generative AI to patent authorities, such as the European Patent Office (EPO) or the Unified Patent Court (UPC), unless specifically mandated by statutes, rules, or client instructions. However, it is essential to stay updated on evolving regulations, as future requirements may change this stance.
4. Are patent attorneys required to disclose the use of AI to clients?
In contrast, when it comes to clients, disclosure is mandatory. Guideline 4 states that patent attorneys must establish the wishes of their clients with regard to the use of generative AI. It is recommended to keep records of enquiries sent to clients in order to establish their preferences with regard to generative AI, and the responses to those enquiries.
5. Are patent attorneys required to seek client approval on the use of AI?
The guidelines do not explicitly state that client approval is required for using generative AI in patent work. Guideline 4 and its explanatory note focus on establishing the “wishes” of the clients. However, it is important to note that the overarching principle in the guidelines emphasizes the importance of putting the client interests first. Therefore, proceeding with generative AI use against a client’s expressed wishes, even if they haven’t explicitly withheld “approval,” would likely be considered a violation of this principle.
Open Questions
Should Clients Be Informed When Purely Local AI Tools Are Used?
The epi Guidelines leave the question of disclosing purely local AI tools to clients open to interpretation. On one hand, Guideline 4 emphasizes the importance of transparency and obtaining the clients’ wishes for the use of generative AI, suggesting that disclosure is necessary regardless of whether the tools are local or cloud-based. Transparency fosters trust, ensures ethical practice, and allows clients to make informed decisions, particularly as they may have concerns about issues such as accuracy or reliance on AI.
On the other hand, the introduction section of the guidelines states that they “may not be applicable” to generative AI tools that amount “to private or confidential systems”, potentially exempting purely local tools if confidentiality and security are fully ensured.
This ambiguity highlights the need for further clarification from the epi regarding whether the use of purely local AI tools always requires disclosure. Until such clarification is provided, attorneys should consider client expectations, ethical standards, and the overarching principles of professionalism when deciding whether to inform clients about local AI use.
Must Patent Attorneys Seek Client Approval for Specific AI Tools?
The epi Guidelines are not entirely explicit on whether patent attorneys must seek client approval for the specific AI tools they intend to use. Guideline 4 requires attorneys to establish their clients’ wishes regarding the use of generative AI, which implies that a general discussion about AI use is mandatory. This could be interpreted to include a need for approval of the specific tools, particularly if those tools raise distinct confidentiality or functionality concerns. Seeking approval for specific tools fosters transparency, ensures alignment with client expectations, and mitigates potential disputes regarding data handling or risk tolerance.
However, the guidelines do not explicitly state that each individual tool requires client approval. Attorneys might reasonably argue that as long as clients are informed about the use of AI in general, the responsibility for selecting appropriate, secure, and compliant tools lies with the attorney. The introduction’s reference to private or confidential systems further suggests that if a chosen tool adequately safeguards confidentiality and aligns with professional standards, explicit approval for that tool might not always be necessary.
This question underscores the importance of further clarification from the epi. Until then, attorneys should consider the nature of the tools, the sensitivity of client data, and the broader principles of trust and client-centered practice when deciding whether to seek specific tool approvals.
Conclusion: Time for Every Patent Attorney to Level Up
The new epi guidelines mark an important step in defining the responsible use of generative AI in patent work. Guideline 1 emphasizes the importance of understanding both the general characteristics of generative AI models and the specific features of the tools we choose to use. Master these essentials in the course “Confidential Patent Drafting with Generative AI: A Beginner’s Guide for Patent Professionals“.
For more insights and updates on using AI responsibly in patent practice, subscribe to the Powerclaim Newsletter for your biweekly dose of AI patent drafting wisdom and a special discount on all Powerclaim Courses: