Microsoft removes ‘AI chatbot’ that hacked 12,000-plus emails in 10,000 firms globally

Microsoft has made significant strides by taking down EvilTokens, a dangerous subscription service that exploited over 12,000 accounts via advanced AI tools. By automating hacking tasks and generating realistic phishing messages from victims’ emails, the platform posed a substantial threat. Following the incident, UK authorities apprehended two individuals tied to the scheme, with Microsoft securing multiple associated domains—illustrating a worrying trend that businesses need to urgently confront.