US Patent Office Bans In-House GenAI Use

The U.S. Patent and Trademark Office has reportedly banned staff from using generative artificial intelligence (GenAI).

The ban has been in place since last April, with the patent office citing security concerns along with the tendency of some tools to display “bias, unpredictability and malicious behavior,” Wired reported Tuesday (Nov. 19), citing an internal memo.

Jamie Holcombe, the chief information officer of the U.S. Patent and Trademark Office (USPTO), wrote that the agency is “committed to pursuing innovation” while still “working to bring these capabilities to the office in a responsible way.”

Paul Fucito, USPTO press secretary, clarified that employees can use “state-of-the-art generative AI models” at work, but not within the office’ internal testing environment.

“Innovators from across the USPTO are now using the AI Lab to better understand generative AI’s capabilities and limitations and to prototype AI-powered solutions to critical business needs,” Fucito wrote in an email.

Outside of the testing environment, USPTO staff are prohibited from using AI programs like OpenAI’s ChatGPT or Anthropic’s Claude to carry out work tasks.

Last year’s memo also forbids workers from using any outputs from these tools, such as AI-generated images or videos. However, USPTO staff can use some approved AI programs, like those within the agency’s public database for looking up registered patents and patent applications, the report said.

This news places USPTO in the company of several other federal agencies trying to ensure the proper use of AI.

For example, PYMNTS reported last month on a meeting convened by the Justice Department’s Civil Rights Division to promote the coordination of AI and civil rights. Experts at that gathering said that auditing can play a role in preventing, investigating, monitoring and remedying algorithmic bias in AI.

“Auditing is used to verify that algorithms generate accurate results, as opposed to reflecting historical bias against protected classes,” a readout from the meeting said.

Also Tuesday, PYMNTS wrote about the concerns that America’s outdated power grid may not be able to handle growing AI usage.

“If the U.S. only relies on traditional approaches to expand the country’s power infrastructure to meet AI infrastructure’s voracious appetite for power, the U.S. will fall behind on its plans to lead in AI globally,” Allan Schurr, chief commercial officer at energy transition company Enchanted Rock, told PYMNTS. “Delays in adding new transmission and generation capacity, which can range from three to 10-plus years, will force companies to seek out alternative locations to support AI infrastructure.”