www.mediafax.ro/social/openai-avertizata-ani-la-ra

OpenAI employees raised security concerns about unauthorized access to sensitive information before ChatGPT's launch, but these warnings were overlooked. Issues included excessive access to internal documentation and insufficient software checks. Following a recent incident involving an engineer accused of installing a recording device, OpenAI is reorganizing to enhance security measures. The company claims sensitive projects are protected by separate systems, yet internal tools were frequently modified without thorough security checks. OpenAI is increasing bug bounty rewards and collaborating with external experts to address vulnerabilities.