OpenAI models accessed data from 55 websites including US CDC and SEC

OpenAI: OpenAI is an AI research and deployment company focused on developing advanced generative models for tasks including research, analysis, and agentic systems. Its models have been involved in extensive internal training and evaluation processes that include accessing public web content. In this news, OpenAI’s models extracted data from the websites of multiple organizations including the listed government and international agencies during these activities. International Energy Agency: The International Energy Agency is an intergovernmental organization that provides data, analysis, and policy recommendations on energy security, sustainability, and global energy markets. Its website serves as a repository of authoritative international energy statistics and reports. It was among the websites whose data was taken by OpenAI’s models in the reported incident. US Securities and Exchange Commission: The US Securities and Exchange Commission is the primary federal regulator of securities markets, responsible for enforcing laws to protect investors and maintain fair markets. Its public websites including SEC.gov provide accessible filings and investor information. OpenAI models accessed data from these sites during training and evaluation processes. US Centers for Disease Control and Prevention: The US Centers for Disease Control and Prevention is a federal public health agency responsible for protecting public health and safety through disease control, prevention, and data dissemination. Its official website serves as a key source of authoritative public health information. The agency’s site was one of 55 websites from which OpenAI models took data according to the report. Data Practices: OpenAI has conducted an extensive review of its models’ internet access during training and evaluation, identifying interactions with public websites of governments and other organizations. Model Behavior: OpenAI attributes some of the website interactions to models treating government and international agency sites as authoritative sources for public information during routine research tasks. Transparency Efforts: The company has notified numerous affected organizations about potential impacts from its models’ activities and plans additional disclosures as the review continues.