AI Email Protection - FAQ

What is the Information Security Team AI Team Agent?

The Information Security Team implemented a product named Abnormal.ai in 2026 in response to the increasing trend of receiving phishing emails generated using AI tools by threat actors starting in 2025. The Information Security Team AI Agent or AI Agent (for short) is the interactive component associated with Abnormal.ai. designed to more effectively filter out increasingly frequent and difficult-to-detect phishing emails and interact with Lamar email users.

How It Works

How does the AI Agent detect threats?
The AI Agent reviews incoming emails for signs of phishing, such as unusual sender behavior, mismatched domains, suspicious links, or unusual writing tone. The Information Security Team regularly reviews the AI Agent's activity and any reported emails.
Is a human also reviewing reported emails?
Yes. Any email reported as phishing is checked by both the AI Agent and a member of the Information Security Team.
Does the system ever make mistakes?
Like any AI tool, the AI Agent is not perfect. Your judgement is still important, and the security team reviews reports to correct any errors.

Using the Agent

How do I report a suspicious email?
Use the "Report Phishing" button in Outlook (Desktop and web versions).

Instructions for Outlook for mobile devices is found here : How do I report phishing or junk email? - Microsoft Support
What happens after I report an email?
You will receive a message from the AI Agent at emailprotection@infosec.lamar.edu You may reply directly to the AI Agent if you want to share more details.
The AI Agent said the email was safe, but I know it was not. What should I do?
Reply to the AI Agent explaining why you believe the email is phishing. Both the AI agent and the Information Security Team will review it again.
Can I ask the AI Agent why it made its decision?
Yes. Reply back to the AI Agent and ask. It will explain its reasoning in plain language.
How long does the AI Agent take to respond?
Usually within a few minutes. If a more extensive review is needed, responses may take longer.

Email Handling and Protection

Will the AI Agent block or delete emails automatically?
Only "high confidence" phishing emails are automatically removed. Emails considered "borderline" or "spam" are sent to your Junk folder.
What if I report an email by accident?
Nothing is blocked automatically. An accidental report does not prevent that sender's emails in the future.
Do I need to delete phishing emails myself?
No. Once an email is confirmed as phishing, it is automatically removed.
What types of threats can the AI Agent detect?
- Business email compromise
- Credential harvesting
- Invoice or payment scams
- AI generated scams
- Malware
- Impersonation attempts
Can I opt out of interacting with the AI Agent?
You can ignore or delete the AI Agent's emails, but reporting helps protect everyone.
Can I opt out of email screening entirely?
No. Email security is essential for protecting the university community.
Who do I contact if I need help?
Email the Information Security Office at infosec@lamar.edu

Data, Privacy and Compliance

Is my email used to train AI models?
No. Your submissions are not used to train any AI model and are not shared outside of LU.
Is Abnormal.AI compliant with FERPA and other regulations?
Yes. The system meets all applicable university, state, and federal standards.
Are my interactions stored?
Your messages with the AI Agent appear in your Outlook "Sent" folder like normal email. Phishing submissions are kept for three months.