OpenAI announced on Thursday that it is beginning the phased release of GPT-6, which the company describes as its most powerful artificial intelligence model to date. The rollout is currently limited to a specific group of customers rather than a general public release.

In its announcement, OpenAI emphasized that the development of GPT-6 includes a significant focus on cybersecurity safeguards. This move comes amid increasing global scrutiny regarding the potential risks posed by advanced AI systems, including their capacity to be misused for cyberattacks or to assist in sophisticated digital threats. By prioritizing these safety features, the company aims to address concerns from regulators and security experts who have warned that more capable models could lower the barrier to entry for malicious actors.

While the company has framed the release as a milestone in AI capability, the limited nature of the rollout suggests a cautious approach to deployment. OpenAI has not yet provided a detailed timeline for when GPT-6 will be available to the broader public, nor have they released specific technical benchmarks comparing the new model's performance to its predecessor, GPT-4o. The current focus remains on testing the model's security protocols within a controlled environment to ensure that the increased power of the system does not inadvertently create new vulnerabilities.