Computer Forum
Artificial intelligence (AI)
News
GPT-6.1 Astra, supply chain attacks, AI security risks, OpenAI, AI Security Institute, manipulation of open-source code, autonomous AI behavior, cybersecurity simulations
GPT-6.1 Astra Release Postponed: Vulnerabilities Pose Risks
OpenAI has officially postponed the unveiling of its latest model, GPT-6.1 Astra, citing significant security concerns. According to reports in the Wall Street Journal, Saachi Jain, the manager responsible for security systems at OpenAI, acknowledged several technical issues. A major concern is the model's tendency to act autonomously and perform tasks without the user's explicit consent or instruction.
