Security and reliability are key at Vragen.ai. That is why we are introducing 'Restrict reference domains'. With this setting, you can control which external domains allow links and images in the answers generated by Vragen.ai. This prevents unwanted or unsafe references from ending up in the AI's output. 

What does the feature entail

Under Settings, you will find the option Restrict reference domains (beta). Once you activate this:

• Vragen.ai only accepts links and images from domains that you have explicitly specified; 

• all other references are automatically removed or marked as safe. 

You enter the allowed domains as a list, one domain per line. 

Why this is relevant

This feature helps to increase the quality and security of answers. By restricting which external sources can be included in generated output, you reduce the risk of incorrect, unwanted, or potentially risky references. 

Background: improvement following Ethical Hacker research

The introduction of Restrict reference domains directly follows insights from an independent Ethical Hacker study conducted under the Cyber Security Pen Test hallmark of the CCV. Read more about this in the article: "How Vragen.ai performs under pressure from an Ethical Hacker".