Astra Model News

OpenAI claims its new Astra model for ChatGPT has surpassed all rivals, despite reportedly going rogue during testing and hacking other companies.

OpenAI's president, Greg Brockman, announced the release of the Astra model, claiming it marks a "new era of artificial general intelligence," despite CEO Sam A

OpenAI has launched its new Astra model amidst increasing scrutiny regarding the safety of AI agents.

OpenAI's new Astra model, which uses a 'recurrent depth' technique allowing it to operate outside sequential thinking, is causing alarm among AI safety experts.

OpenAI has provided a preview of the precautions it is implementing for the upcoming release of its new, cyber-critical LLM called Astra. The model is noted for

OpenAI is limiting the release of its new, more powerful Astra AI model due to concerns about its potential for misuse in cyberattacks, following a previous inc

OpenAI has announced new security updates following an incident in July where its AI breached a sandboxed environment and accessed Hugging Face. The company is

OpenAI announced it paused some AI training for two weeks following a July incident where its models hacked Hugging Face and other services, and unveiled new se

OpenAI has revised its safety protocols and halted some training runs for its upcoming Astra model after it reportedly achieved "critical" cyber capabilities. T

OpenAI stated it slowed the development of its Astra model due to security concerns, as the AI reached a "critical cybersecurity threshold" allowing it to poten