On April 10, 2024, the USA Patent and Commerce Workplace (USPTO) introduced steerage concerning the usage of synthetic intelligence (AI) instruments earlier than the USPTO.¹ The steerage acknowledges that AI instruments include potential advantages—improved prices, high quality, and effectivity—and potential dangers—corresponding to inaccuracy and inadvertent disclosure of confidential data—and focuses on how current guidelines and insurance policies apply within the context of AI instruments.²

Major Takeaways

The first takeaways from the steerage embrace the next, with additional particulars beneath:

1. The USPTO has decided that current guidelines are enough to guard in opposition to the present dangers of utilizing AI instruments.
2. Utilizing AI instruments to help with drafting paperwork for submission to the USPTO will not be prohibited.
3. There isn’t a normal or per se obligation to reveal when an AI device is used when practising earlier than the USPTO or within the invention creation course of, except particularly requested by the USPTO.  Nevertheless, there could also be particular instances the place such disclosure is required.
4. The steerage gives strategies in numerous examples concerning the extent of human involvement or oversight wanted to adjust to current guidelines and insurance policies.

Overview of the Steering

The steerage features a part reviewing current guidelines which may be related to the usage of AI instruments when practising earlier than the USPTO, particularly: the obligation of candor and good religion, signature necessities and corresponding certifications, confidentiality of data, international submitting licenses and export rules, USPTO digital methods insurance policies, and duties owed to shoppers.

The steerage then turns to the applying of these current guidelines to the usage of AI earlier than the USPTO, beginning with its use in doc drafting.  The steerage acknowledges that AI instruments could also be used to draft patent specs or claims, Workplace Motion responses, briefs, affidavits, or petitions.  A practitioner should evaluate any paper submitted to the USPTO for accuracy and to make sure the arguments are legally warranted.  Merely counting on the accuracy of an AI device will not be enough.

As well as, there are circumstances the place the usage of an AI-based device is materials to patentability such that disclosure, together with doubtlessly the inputs/outputs of the AI system, is required.  For instance, if an AI-based device is used to draft a specification or declare, the AI-based device might suggest a number of different embodiments.  In such circumstance, a query arises as as to whether the a number of different embodiments proposed by the AI-based device are entitled to patent safety, i.e., whether or not a human “considerably contributed to a claimed invention developed with the help of AI” such that it’s patentable material.

The steerage additionally cautions the usage of AI instruments to mechanically populate Info Disclosure Assertion (IDS) kinds might improve the quantity or dimension of IDS submissions, risking burdening the USPTO with giant numbers of cumulative and irrelevant submissions.  The steerage states that IDS contents needs to be reviewed to take away clearly irrelevant and marginally pertinent cumulative data, and that submitting an IDS with out reviewing its contents could also be a violation of 37 CFR 11.18(b).

The steerage additionally gives data concerning the usage of AI instruments to work together with USPTO methods, together with submitting paperwork with the USPTO and accessing USPTO databases.  For instance, the steerage explains that AI-based instruments can not signal filings or have a consumer account for accessing USPTO methods—these actions should be carried out by an individual.

The steerage ends by recommending vigilance when disclosing data to AI instruments to keep up confidentiality of shopper information, and to contemplate nationwide safety, export management, and international submitting license points (e.g., if AI instruments use servers positioned exterior the USA), and with a reminder that AI instruments shouldn’t be used for the unauthorized entry, use, or disclosure of information from USPTO methods.

Conclusion

The AI Steering, efficient April 11, 2024, gives useful data concerning the usage of AI-based instruments for USPTO observe.  Practitioners ought to rigorously evaluate the steerage and ensure that their use of AI instruments complies with the USPTO’s guidelines and procedures.