Cisco Talos researchers identified a 150% surge in the weaponization of n8n, a legitimate, open-source AI workflow automation platform, between October 2025 and March 2026. Over 3,000 malicious email campaigns were detected in this six-month window. Attackers used n8n's built-in agentic AI capabilities to dynamically generate malicious payloads, bypass sandbox security filters, and automate both phishing delivery and malware distribution at scale. Because n8n is a trusted, widely recognized tool, many traditional email security filters failed to flag the campaigns, allowing attackers an average dwell time of 18 days before detection. That is nearly three weeks inside a network before anyone noticed.

Ontario SMBs are adopting automation tools at an accelerating pace. Legal firms in Vaughan are automating client intake. Accounting practices in Markham are automating report generation. Construction companies in Brampton are automating supplier communications. Dental clinics in Oakville are automating patient follow-ups. This wave of adoption is smart, but it creates a new attack surface that most businesses are not equipped to defend. When attackers abuse platforms like n8n, they look legitimate to your email gateway, your firewall, and even your employees. The payload arrives wrapped in the appearance of a normal workflow notification. There is no obvious red flag. Your team clicks. The breach begins. For a 15-person accounting firm or a 30-person law office, 18 days of undetected attacker access can mean the complete compromise of client data, financial records, and regulatory standing under Ontario's privacy laws.

Here is how the attack unfolds in plain language. Attackers set up their own instance of n8n, which is free and open-source, and build automated workflows that mimic legitimate business communications. These workflows are programmed to send tailored phishing emails, dynamically swap out malicious payloads to avoid pattern detection, and adapt in real time when a sandbox flags a variant. Because the emails originate from automation logic that resembles real business workflow traffic, traditional spam filters and even some enterprise security tools struggle to identify them as threats. Once an employee clicks a link or opens an attachment, the attacker gains a foothold and uses n8n's agentic AI capabilities to move laterally through the network, accessing shared drives, credentials, and connected cloud services, all while staying under the radar for weeks.