AI-Powered conversational phishing simulations (over both email and voice) can be improved by giving your unique AI Agent access to relevant information about the target organization. Open source intelligence provides a way to collect this information using publicly available data.
OSINT helps the AI generate context-aware conversations. These details are used to make each simulated phishing conversation more realistic and more aligned with real-world attack patterns.
The only information used to conduct the scan is your organization's name and your domain name.
To leverage the power of OSINT, please follow the below steps:
Step 1: Go to Platform Settings > Phish Settings > AI Knowledge Source:
Step 2: Click "Build AI Knowledge Source Using OSINT" and then click "Yes, let's do it!" The process will take up to a minute.
You may need to refresh the page if the knowledge source does not automatically update.
Step 3 (Optional but recommended): Add additional details to the AI Knowledge Source about which collaboration and HR systems you use. This greatly enhances the realism of your voice phishing simulations.
Step 4: Click "Deploy Knowledge Source" and click "Yes, publish it"
The status will change from "Unpublished" to "Published"
Frequently Asked Questions
Can I change the knowledge source?
No. Scan results are read-only and cannot be modified or otherwise altered - you can however adjust the collaboration and HR systems we detect.
What kind of data does the AI use to personalize phishing exercises?
The AI uses any data provided in the knowledge source plus the employee information provided in the employee list.
Where does this data come from?
The data can be collected from public sources such as company websites, social media profiles, press releases, or breached data repositories—anything openly available without requiring authentication.
Can I opt out of OSINT-based personalization?
Yes. OSINT-based personalization is turned off by default and requires explicit activation. It can also be deleted/disabled at any time.
Comments
0 comments
Article is closed for comments.