OpenAI lanza GPT-6 Astra, su IA más potente puede operar una computadora de forma autónoma
2026-09-04
El Universal
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
OpenAI has begun deploying GPT-6 Astra, its most powerful AI model, capable of autonomously operating computers and hacking well-protected systems. The launch follows recent AI-driven cyberattacks involving earlier OpenAI tools, raising significant cybersecurity concerns. OpenAI is implementing risk mitigation and initially restricting access to select organizations.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly mentions an AI system (GPT-6 Astra) with autonomous capabilities including hacking, which is a direct cybersecurity risk. While no actual harm or incident is reported from GPT-6 Astra's deployment yet, the model is described as capable of causing serious harm (e.g., hacking well-protected systems) and previous AI tools have been involved in cyberattacks. The deployment includes risk mitigation and restricted access, indicating awareness of potential dangers. Since the harm is plausible but not yet realized, this fits the definition of an AI Hazard. The article does not focus on a realized harm or incident, nor is it primarily about responses or updates to past incidents, so it is not Complementary Information. It is not unrelated or beneficial use, as the AI system's capabilities pose a credible risk of harm.[AI generated]