Catch the Recording of AI Threats All M365 Admins Should Know

March 11, 2024
2 min read
Example of AI generated fake email
Example of convincing generated fake email View Full Size

I hinted in my previous story that having your company services halted by a ransomware attack is in all ways a tough experience for everyone. There are a lot of ways to mitigate that from happening, but probably the most effective is having really good commercial software that scans all of your inbound corporate email and have it intercept ransomware delivery email before anyone clicks the bad link.

Fast forward a few years and I find myself wearing the hat of being an occasional Microsoft 365 Admin, and there’s a whole lot of fear, uncertainty, and doubt about how good or bad artificial intelligence (AI) is going to make that part of my job. Are you at a company that trusts Microsoft 365 will just handle it all for you if your tenant gets a ransomware attack? Some credible sources I’ve read suggest that while being in the cloud lowers the risk a bit, Microsoft doesn’t automatically throw you a life preserver and stuff just works out in your favor. It will pay off, eventually, to take more steps now to bolster your security.

On March 7, 2024, Christian Buckley MVP, RD hosted a webinar with Dr. Kiri Addison and J. Peter Bruzzese named “AI Threats All M365 Admins Should Know” on TekkiGurus. In a fun exchange at the beginning, Kiri stumped J. Peter a few times trying to identify whether some recent emails received by customers were created by a person or AI. A lot of the regular “tells” that we’ve looked for in years past (bad grammar, poor spelling, awkward language) aren’t necessarily in some of those email messages anymore. AI is helping hackers write more convincing emails.

AI can also help the good guys. You likely work at a company that has received millions of emails, and within those, a few hundred or a few thousand of them might be “bad emails” that could subject your company to a ransomware attack. Cybersecurity company Mimecast trains their email classification AI models on the 26 billion monthly email records they scan for their customers worldwide. They have more examples of “bad emails” to train their AI models on. Using their tools, they will likely identify more of the bad messages coming in than anything your company is using.

I encourage you to watch the webinar that TekkiGurus presented with Mimecast. You will absolutely learn things to improve the security of your Microsoft 365 tenant.

Erik Ruthruff

Erik Ruthruff

Erik is Editor-in-Chief of TekkiGurus. He has worked in computer education for software professionals since the mid-90s through editing, training, conferences and project management. A Certified Professional Project Manager (CPPM), Erik also works at CODE as a project manager and helps with CODE Magazine. He has been a member of PMI Minnesota since 2017. In his spare time, he cycles, plays D&D and does non-profit work. Erik spent ten years as a volunteer for the A Better Chance foundation, three years as a volunteer commissioner for the Edina Transportation Commission, and now volunteers for as a content editor and member of the executive board.