OpenAI has announced that it will do something very special to protect its users from copyright issues.
The introduction of a new scheme, entitled Copyright Shield, will see the company foot any bills relating to GPT copyright cases.
There’s a catch, though, because Copyright Shield only applies to customers using generally available products, which could leave many regular consumers in limbo.
Using ChatGPT won’t cost you in any legal battles
In an announcement, the company stated: “OpenAI is committed to protecting our customers with built-in copyright safeguards in our systems.”
OpenAI added: “we will now step in and defend our customers, and pay the costs incurred, if you face legal claims around copyright infringement.”
Copyright Shield only protects users running generally available OpenAI products, including ChatGPT Enterprise and the developer platform.
The news comes after months of pressure – Microsoft announced that it would protect its own customers against legal battles by footing the bill earlier in September, two months before OpenAI’s move.
Copyright in the world of generative AI has become an increasingly important topic, because models like the firm’s latest GPT-4 use publicly available information for training purposes. This means that, on occasion, AI-generated responses could include content that belongs to somebody else.
Copyright protection is even more important in the case of OpenAI, which, at the same event, announced that it would make its sources much more current. The GPT-4 Turbo model now has knowledge of real-world events up to April 2023.
It’s worth noting that, at least for a few more weeks, GPT-4 turbo is in preview, so Copyright Shield won’t actually step in in the event of an infringement allegation.
Clearly, protection against copyright cases for products that are still in development would be a foolish move on OpenAI’s part, but its offer for generally available products demonstrates a clear commitment to responsible AI, just days after a British-organized summit saw representatives from various countries and enterprises discuss the future of AI.
More from TechRadar Pro
- 28 countries agree to develop AI safely and responsibly at Bletchley Park summit
- Want to help boost productivity at work? Here are the best AI tools
- Craft your best work every with the best AI writers
Millions of devices still connect to this dangerous malware, despite the creators ditching it years ago
Finally! After a 7-year wait, this monitor could well be the best pro-level 8K display ever — will Asus be able to break the curse of failed 8K monitor launches with the PA32KCX Mini LED pro screen?
