AI NewsWaveOpenAI, Google and Microsoft Back Common Standard for Opting Out of AI Training
Key points
- OpenAI, Google, Microsoft and others back a common standard for opting out of AI training
- Website operators can refuse collection all at once
- Shift from blocking each company's crawler bot separately
Major AI companies including OpenAI, Google and Microsoft have decided to support a common standard that lets website operators opt out of AI training data collection all at once. Chosun Biz reported that this was announced on the 31st (local time), and IT Chosun reported that on the same day the three companies would let operators refuse AI training content collection all at once.
According to Chosun Biz, operators until now had to block each company's crawler bot one by one in a 'robots.txt' file. Digital Today, citing The Verge, reported that website operators had to tell each AI company separately whether to allow their data to be used for AI training. AI Newspaper (인공지능신문) also explained that the standard moves away from the approach of setting a separate crawler access policy for each AI company.
Reports differed on which companies are taking part. Digital Today named four companies, adding Anthropic to OpenAI, Google and Microsoft, while Chosun Biz wrote 'OpenAI, Google, Microsoft and others.' In its afternoon article on the 31st, AI Newspaper (인공지능신문) reported that technical standardization is being 'pursued' in which content publishers mark directly whether their content may be used for training and multiple AI companies recognize this in common, which differs in wording from other outlets' 'adoption.' IT Chosun added that protection of rights over online content is expected to be strengthened further.
Glossary robots.txt is a text file that tells automated collection programs, such as search robots, which parts of a website they may access. A crawler is a program that automatically goes through web pages and collects their content.