Stop copying and pasting from ChatGPT as new feature can actually tag everything you write
Anyone who has been sneakily copying and pasting text from ChatGPT may want to heed this warning from OpenAI as the firm announces it will start introducing an invisible watermark.
The AI giant shared the update in a blog post on Monday (October 5), which revealed how OpenAI is responding to new rules laid out in the EU AI Act.
OpenAI explained: “The EU AI Act requires generative AI providers to make generated text identifiable in a machine-readable way. Text watermarking and detection remain early technologies with significant limitations, and views about their benefits and responsible uses are still developing.
“Our phased approach reflects both the EU AI Act requirements as well as the technology’s limitations, with an emphasis on transparency about what a text watermark can and cannot tell people.”
Advert
From this week, API customers around the world will be able to opt in for text watermarking for certain AI models, although the watermarking will be off by default in the API.
In the coming weeks, OpenAI will ‘add an invisible watermark to eligible ChatGPT and Codex text output in the European Union’.
Applications are being opened for access to the firm’s ‘text watermark detector’, with access initially being limited to approved researchers and expert organizations ‘that can help us evaluate and improve the technology’.
How does the ChatGPT watermark work?
The watermark won’t be a physical feature like a symbol within the text, and instead it will be a pattern in the way ChatGPT writes which will reveal itself to a detector.
OpenAI continued: “Our text watermarking technology, textGrain, adds an invisible statistical signal to the model’s word choices. Our detector looks for that signal to assess whether a passage contains an OpenAI watermark.
“More details about how textGrain works can be found in our technical report, which will be updated with additional details in the coming weeks. We also plan to make the technology available in open source so that others can build on it.”
EU cracks down on AI
In August, the EU implemented new powers to allow its executive arm, known as the Commission, to inspect models being introduced to the market under its new AI Act.
AI models put on the market will have until August 2, 2027, to comply with new obligations.
The Council of the European Union described the act as being a way to ensure that AI systems are ‘developed and used responsibly’.
It stated: “The rules impose obligations on providers and deployers of AI technologies and regulate the authorisation of artificial intelligence systems in the EU single market.
“The law addresses risks linked to AI, such as bias, discrimination and accountability gaps, promotes innovation and encourages the uptake of AI.
“As the world’s first law regulating AI, the EU’s rules could set a global standard in AI regulation, just as the general data protection regulation (GDPR) has done for data privacy, promoting ethical, safe, and trustworthy artificial intelligence worldwide.”