AI-related security incidents have been spreading well beyond email and into the wider collaboration stack. The data, coming from the latest Proofpoint survey, analyzed the responses of 1,453 security professionals across 12 countries.

Among the 42% of respondents whose organization had experienced an AI-related incident, threat activity was reported in email (67%), SaaS or cloud apps (57%), AI assistants or agents (53%), and collaboration tools, social platforms, and file-sharing (49% each). Email remains the most common entry point, but the gap between it and the next channels has narrowed sharply.

The pattern tracks with how AI is being deployed inside enterprises. According to the report, 69% of respondents said their organization uses AI assistants for customer or technical support, 67% for chat summarization in Slack or Teams, and 63% for email drafting. The channels carrying the most AI-mediated work are also the ones where threats are landing.

“While AI has introduced new risks, such as prompt engineering, its bigger impact has been amplifying the risks we’ve always had,” said Ryan Kalember, Chief Strategy Officer at Proofpoint. “Running untrusted code, mishandling sensitive data, and losing control of credentials are the same challenges that humans have created for decades.”

Cross-Channel Investigation Is the Weak Point

Detection is one problem, reconstructing what happened is another, the report explained.

Only 33% of respondents said they were fully prepared to investigate an AI- or agent-related incident, and 41% reported difficulty correlating threats across multiple channels. Tool sprawl is a meaningful contributor, with 95% describing the management of multiple security tools as at least moderately challenging.

The findings arrive months after Proofpoint’s strategic pivot toward what the vendor calls the “collaboration firewall,” a platform position that assumes attacks will keep spreading off email and into the channels where work actually happens.

Looking ahead, 61% of respondents plan to expand AI protections over the next 12 months, 56% intend to extend coverage across collaboration channels, and 53% expect to move toward a unified platform.

The survey was conducted in January 2026 and covered respondents in the US, the UK, France, Germany, Italy, Spain, the UAE, Australia, Brazil, India, Japan, and Singapore.