The US government’s order targets OpenAI’s AI model. The ChatGPT developer OpenAI is restricting access to its latest AI model at the request of the US government. Only a small, vetted group of trusted partners will receive access to the preview version of models in the GPT-5.6 series, OpenAI said. The limited release was demanded by the government.
Authorizations have so far only been granted in the US, OpenAI said. The company is working to give partners from other countries access as well, which could happen as early as next week.
OpenAI also stressed that it does not believe such government access controls should become the norm long-term. For now, it sees agreeing to them as the best way to expand access in the coming weeks.
This statement can be seen as a reference to problems faced by rival Anthropic. The company initially released its new, most powerful model – only to have to withdraw it a few days later at the request of the US government. The trigger was warnings that safeguards against misuse of the software could potentially be bypassed. Anthropic’s new AI model is particularly adept at finding vulnerabilities in software. That makes programs more secure – but in the wrong hands, such an AI could be used as a cyber weapon. Anthropic blocks those capabilities in the publicly available version. However, the US government demanded that only American companies and individuals be allowed access to the AI model.
With OpenAI’s new GPT-5.6 model, the most powerful variant, named Sol, can independently perform tasks in programming, biology, and cybersecurity, according to the company. OpenAI emphasized that it has implemented safeguards against misuse. At the same time, the company warned that evaluations cannot account for every multi-step attack.
Source: www.tagesspiegel.de
