OpenAI has notified dozens of international institutions that its AI tools improperly extracted data from their websites, BBC reports.
The tools pulled data from public websites belonging to government agencies and universities, among others. In at least 53 cases, the AI tool took images users had uploaded to ChatGPT and forwarded them elsewhere.
OpenAI has apologized and called the behavior inappropriate. The company is now working to delete the data that was improperly obtained. The notification comes just days after it emerged that OpenAI’s AI had accessed the Hugging Face platform without authorization over the summer.
