Free daily briefing on global business news.
Preliminary evaluations found the unreleased model may have reached a "critical" cybersecurity threshold under the company's safety framework
SOPA Images / Getty Images
OpenAI said Friday it has paused some internal activities involving its upcoming model Astra after preliminary evaluations found it may be capable of independently launching cyberattacks against well-protected systems — capabilities that triggered additional safety protocols under the company's Preparedness Framework.
That framework, introduced in 2023, designates a model as "critical" when it demonstrates the ability to independently find and exploit severe software vulnerabilities in real-world systems, or carry out sophisticated cyberattacks on heavily secured targets without any human direction, according to Reuters. "While we continue to benchmark and assess this model, our preliminary evaluations indicate strong enough performance that we cannot rule out Critical capability level at this time," the company said in a statement.
