Threat
Invisible Unicode Tags Hide Prompts From AI Email Agents
First reported · Discovered securitypointbreak.com
Page published
Earliest dated coverage: 28 Oct 2026 · First observed: 4 Nov 2026 · Latest dated coverage: 28 Oct 2026
Coverage timeline
Single-source incident — one report is available.
Why it matters
Invisible Unicode prompt injection exploits differences between human-visible email and machine-parsed content, making visual inspection insufficient to identify instructions targeting inbox agents.
Attackers are using invisible Unicode tag characters in fake-invoice emails to evade keyword filters and conceal instructions aimed at AI email agents. Sublime Security researcher Luke Wescott reported observing real-world prompt-injection attempts using the technique, but said those attempts did not succeed.