BRUSSELS, Belgium: EU tech chief Margrethe Vestager has said that a draft code of conduct on artificial intelligence (AI) could be drawn up within weeks, so the computer industry can commit to a final proposal "very, very soon," as reported by Reuters.

Comparing it to the risks posed by pandemics or nuclear war, some EU lawmakers and many in the industry have expressed concern about AI, most notably content-creating generative AI, such as ChatGPT.

The US and the EU should push a voluntary code to provide safeguards while new laws are developed, Vestager said, stressing, "Generative AI is a complete game-changer."

"Everyone knows this is the next powerful thing. So within the next weeks we will advance a draft of an AI code of conduct," she said, adding she hoped there would be a final proposal "very, very soon" that industry could sign up to, according to Reuters.

Watermarking, external audits and other controls could feature in the code, she later tweeted.

The EU's AI Act, which regulates facial recognition and biometric surveillance, is still passing through the legislative process.

"In the best of cases, it will take effect in two and a half to three years time. That is obviously way too late. We need to act now," Vestager told reporters in Sweden.

This month, G7 leaders called for the adoption of technical standards to keep AI "trustworthy," as well as international discussions on related topics, such as governance, copyright, transparency and disinformation.