OpenAI releases latest ChatGPT model after delay over cybersecurity concerns

OpenAI has finally released its new advanced AI model, named ChatGPT 5.6. The release was delayed because the US government had worries about cybersecurity. Last month, the Trump administration asked OpenAI to only allow a small number of government-approved users to access the new model.

OpenAI agreed to this request from the White House. The company explained that it had informed government officials about what ChatGPT 5.6 could do and limited its use to trusted partners as requested. The model was made available to more people after the government's Center for AI Standards and Innovation agency conducted further tests, according to reports.

This controlled release of ChatGPT 5.6 is similar to the restrictions placed on the newest AI models from a competing company, Anthropic. The Trump administration also put limits on Anthropic's models last month, which led to a temporary ban on exporting them. These events highlight the cybersecurity dangers linked to AI, disagreements with other countries about sharing new technology, and questions about the US government's quick and sometimes messy way of creating rules for this technology.

OpenAI says that its new model, ChatGPT 5.6, which includes a main product called Sol, is its safest and most capable one yet. It is designed to compete with Anthropic's Claude Fable and Mythos models. These models were released to the public recently after facing similar government limitations.

The Trump administration has generally ignored calls from within the US to slow down the AI industry. Instead, it has encouraged companies to develop their AI models quickly, often mentioning a competition with China and the importance of the US leading in technology. Recently, however, the risk of serious cyber-attacks caused by AI has made the White House more involved. Donald Trump signed an order last month that suggested AI companies should voluntarily submit their new models to the government for review.

The US government's decision to control who can use the most advanced AI models has created worry for businesses that depend on these tools. It has also concerned other countries that might face cybersecurity threats and allies who felt overlooked by the US decision to ban exports. Initially, OpenAI only allowed US companies to access its newest model.

Both Anthropic and OpenAI are competing to be the leader in AI before they plan to offer their company shares to the public later this year. In May, Anthropic announced it had received new investments, making it the most valuable AI company in the world. OpenAI is reportedly trying to achieve an even higher valuation for its own stock market debut, though reports suggest this might be delayed until next year due to instability in the global technology market.