How to Ask ChatGPT and Google Bard to Not Use Your Website for Training

Dec 12, 2023 ·

Recent changes to the Robots.txt protocol have made it impossible for websites to ask search engines and language models not to use their data for training purposes. This has caused a great deal of concern in the web development community, as companies rely on their data being protected from automated and algorithm-driven harvesting. However, this is becoming increasingly difficult, with Google, OpenAI, and other big players all using web content to train their artificial intelligence systems.

The EFF recently analyzed these changes and proposed a new protocol called NoRobotsTXT. It would allow web developers to protect their data by asking search engines and language models not to use it for training purposes. The protocol would require an explicit request from the website owner, rather than relying on the robots.txt file.

The EFF's proposal raises some interesting questions about how AI systems should be developed. On the one hand, it could lead to more responsible AI development, as companies would need to consider the implications of using web content for training. On the other hand, it could prevent companies from taking advantage of existing web content to build better AI systems.

The debate around NoRobotsTXT will likely continue for some time, as both sides make their case. In the meantime, it is important for web developers to know their rights when it comes to protecting their data, and to understand the implications of allowing search engines and language models to access it. As technology continues to develop, it is essential that companies take responsibility for the data they use and respect the privacy and security of those who provide it.