OpenAI Pauses Work on Next Model as ‘Critical’ Cyber Capabilities Raise Global Security Fears
OpenAI has halted some work on its next AI model, codenamed Astra, after internal testing suggested it might enable “critical” cyber operations beyond what current safeguards can manage. The move exposes how frontier AI development is colliding with national security, as governments and attackers race to exploit or contain new digital capabilities.
One of the world’s leading AI labs has quietly hit the brakes on part of its most advanced project, after internal tests suggested the new system could unlock “critical” cyber capabilities that raise alarms far beyond Silicon Valley. OpenAI has paused some work involving Astra, its next‑generation AI model, and tightened internal security controls, according to a 10 August report in the cybersecurity press. The decision follows internal evaluations indicating that Astra could assist in cyber operations at a level that OpenAI is not yet confident it can safely contain or monitor. The company has not publicly disclosed technical details of the capabilities or specific scenarios that triggered concern. For…
Pro features include
- 60+ analytical tools across markets and intelligence
- Custom alerts, watchlists, and AOI monitoring
- Daily Pro brief at 6 PM ET — 12 hours before free tier
- Conflict deep dives and premium research products