By Supantha Mukherjee

STOCKHOLM (Reuters) -Microsoft ( MSFT ) is offering free of charge to European governments a cybersecurity programme, launched on Wednesday, to bolster their defences against cyber threats, including those enhanced by artificial intelligence, it said.

After a surge in cyberattacks in Europe, many linked to state-sponsored actors from China, Iran, North Korea and Russia, the programme aims to boost intelligence-sharing on AI-based threats and help to prevent and disrupt attacks.

"If we can bring more to Europe of what we have developed in the United States, that will strengthen cybersecurity protection for more European institutions," Microsoft ( MSFT ) President Brad Smith told Reuters in an interview.

"You're going to see other things we are doing later in the month."

Increasingly, attackers employ generative AI to amplify the scale and impact of their operations that range from disrupting critical infrastructure to spreading disinformation.

Although malicious actors have weaponised AI, Smith said AI also offered defensive tools.

"We don't feel that we have seen AI that has evaded our ability to detect the use of AI or the threats more broadly," Smith said.

"Our goal needs to be to keep AI advancing as a defensive tool faster than it advances as an offensive weapon," he said.

Microsoft ( MSFT ) tracks any malicious use of AI models it releases and prevents known cybercriminals from using its AI products.

AI-driven deepfakes have included a portrayal of Ukrainian President Volodymyr Zelenskiy capitulating to Russian demands in 2022 and a fake audio recording in 2023 that influenced the Slovakian election.

Smith said so far audio had been easier to fake than video.