Looks like the US Patent and Trademark Office (USPTO) has banned almost all use of generative AI. Employees can still try it out in controlled experiments, but not for anything serious. They’re worried about security, bias, and AI making unpredictable or even malicious decisions. Thoughts?
Welcome to the AI forum
Some reminders for posting news
Make sure your post follows these guidelines:
- Posts need to have enough detail to start a discussion. Avoid super short posts.
- Always share a direct link to the article or source.
- If you’re sharing a personal blog or project, let people know your connection.
- Add a short description of the news to help others understand what it’s about.
- There’s a lot of AI news out there, so engage people with your thoughts.
Let the moderators know if you have any questions!
I am an automated bot. Reach out to the moderators if you have feedback or need help.
The USPTO banned generative AI because of security issues and concerns about bias and unpredictable results. Their chief information officer said they want to innovate but need to do it responsibly. Full details here: https://www.wired.com/story/us-patent-trademark-office-internally-banned-generative-ai/
This makes sense. Most generative AI tools openly admit that anything you give them could be used to train their models. If you don’t see how that could clash with the goals of a patent office, I don’t know what else to say. It seems like a big risk for them to take.
Here’s one way to think about generative AI:
These tools create a ton of options and then pick what they think is the “best” result. But they don’t know everything and can’t predict the future, so sometimes their answers are wrong or outdated. Relying on it for critical tasks might not be the smartest idea.
What’s everyone else’s take on this?
The office meant to protect ideas doesn’t want to use a tool that takes other people’s ideas—makes sense. Most AI models say they might use what you give them to train their systems. That doesn’t really align with what a patent office does. Plus, companies like Coca-Cola don’t even patent some inventions to avoid making secrets public. Remember when ChatGPT used to spit out things like bomb instructions? Yeah, that’s a concern.
Feels like they’re just playing it safe to keep their jobs secure.