President Donald Trump has delayed an executive order aimed at ensuring the security of artificial intelligence models, citing dissatisfaction with the language and concern that it might hinder America’s global lead.

The order was expected to task relevant government agencies with developing a process for evaluating AI models before their release. This comes in response to concerns over recent AI releases from Anthropic and OpenAI, both of which could potentially exploit security vulnerabilities.

Trump’s decision has been attributed to the difficulty in scheduling tech CEOs for a photo opportunity in Washington D.C., according to multiple reports. While the president expressed his commitment to leading China and other nations in technology, he seems to prioritised this over the potential risks posed by unregulated AI.

The sticking point identified was a proposed requirement for companies to share their advanced models with government officials ahead of release, with a timeframe of 14 to 90 days. Trump expressed concern that this could have hindered the order’s effectiveness.