The New York Times reported that OpenAI transcribed more than a million hours of YouTube videos to train its AI model called GPT-4, acknowledging the legal concerns but considering it to be fair use.
"OpenAI president Greg Brockman was personally involved in collecting videos that were used," according to the report.
An OpenAI spokesperson mentioned that the company leverages 'numerous sources including publicly available data and partnerships for non-public data' to ensure its competitiveness in global AI research.
Google, the owner of YouTube, stated it had 'seen unconfirmed reports' about OpenAI's actions, emphasizing that unauthorized scraping or downloading of YouTube content is prohibited by their robots.txt files and Terms of Service.
[
add
]
[
|
|
...
]