NEW YORK: OpenAI, the renowned AI research organization based in San Francisco, faced a significant security breach last year when a hacker infiltrated its internal messaging systems and accessed details about the company’s artificial intelligence technologies, the New York Times reported.
According to the report, the hacker managed to extract information from discussions held in an online forum where OpenAI employees discussed the organization’s latest advancements in AI technology. However, the breach did not extend to the systems where OpenAI houses and develops its AI models, including the popular ChatGPT chatbot.
Sources familiar with the incident, cited in the report, disclosed that OpenAI executives reported the breach during an all-hands meeting in April last year, informing both employees and the company’s board. Despite these internal communications, OpenAI opted not to publicly disclose the breach, as no sensitive information related to customers or partners was compromised.
The breach was not considered a national security threat by OpenAI executives, who assessed the hacker as an individual without apparent ties to any foreign government, the report added. Consequently, the company did not report the incident to federal law enforcement agencies.
Earlier this year, OpenAI announced that it had disrupted five covert operations aimed at exploiting its AI models for deceptive activities online.
Meanwhile, as per a previous Reuters report, the Biden administration was reportedly preparing to implement new measures to protect US AI technology from threats posed by countries like China and Russia. These measures include introducing safeguards specifically targeting advanced AI models, including ChatGPT.
In a global meeting in May, sixteen leading AI companies pledged to prioritize the safe development and deployment of AI technologies amid growing regulatory challenges and emerging risks.
