The US Patent and Trademark Office Banned Staff From Using Generative AI

The US Patent and Trademark Office banned the usage of generative artificial intelligence for immoderate intent past year, citing information concerns with the exertion arsenic good arsenic the propensity of immoderate tools to grounds “bias, unpredictability, and malicious behavior,” according to an April 2023 interior guidance memo obtained by WIRED done a nationalist records request. Jamie Holcombe, the main accusation serviceman of the USPTO, wrote that the bureau is “committed to pursuing innovation wrong our agency” but are inactive “working to bring these capabilities to the bureau successful a liable way.”

Paul Fucito, property caput for the USPTO, clarified to WIRED that employees tin usage “state-of-the-art generative AI models” astatine work—but lone wrong the agency’s interior investigating environment. “Innovators from crossed the USPTO are present utilizing the AI Lab to amended recognize generative AI's capabilities and limitations and to prototype AI-powered solutions to captious concern needs,” Fucito wrote successful an email.

Outside of the investigating environment, USPTO unit are barred from relying connected AI programs similar OpenAI’s ChatGPT oregon Anthropic’s Claude for enactment tasks. The guidance memo from past twelvemonth besides prohibits the usage of immoderate outputs from the tools, including images and videos generated by AI. But Patent Office employees tin usage immoderate approved AI programs, such arsenic those wrong the agency’s ain nationalist database for looking up registered patents and patent applications. Earlier this year, the USPTO approved a $75 cardinal contract with Accenture Federal Services to update its patent database with enhanced AI-powered hunt features.

The US Patent and Trademark Office, an bureau wrong the Department of Commerce, is successful complaint of protecting inventors, awarding patents, and registering trademarks. It besides “advises the president of the United States, the caput of commerce, and US authorities agencies connected intelligence spot (IP) policy, protection, and enforcement,” according to the USPTO’s website.

At a Google-sponsored lawsuit successful 2023, Holcombe, the writer of the guidance memo, said government bureaucracy makes it hard for the nationalist assemblage to usage caller technologies. “Everything we bash successful the authorities is beauteous stupid, erstwhile you comparison it to the commercialized world, right?” helium said. Holcombe specifically cited cumbersome budgeting, procurement, and compliance processes, arguing that they hamper the government's quality to rapidly follow innovations similar artificial intelligence.

The USPTO is not the lone authorities bureau to prohibition unit from utilizing generative AI, astatine slightest for immoderate purposes. Earlier this year, the National Archives and Records Administration prohibited the usage of ChatGPT connected government-issued laptops, according to 404 Media. But soon afterward, the National Archives hosted an interior presumption that encouraged employees to “think of [Google’s] Gemini arsenic a co-worker.” During the meeting, immoderate archivists reportedly expressed concerns astir the accuracy of generative AI. Next month, the National Archives is readying to merchandise a caller nationalist chatbot for accessing archival records developed with exertion from Google.

Other US authorities agencies are using—or avoiding—generative AI successful antithetic ways. The National Aeronautics and Space Administration, for example, specifically banned the usage of AI chatbots for delicate data. NASA did decide, however, to experiment with the exertion for penning codification and summarizing research. The bureau besides announced past week that it’s moving with Microsoft connected an AI chatbot that tin aggregate outer data to marque it easy searchable. That instrumentality is disposable lone to NASA scientists and researchers, but the extremity is to “democratize entree to spaceborne data.”