AI may be unable to access a server due to network restrictions, proxy or firewall settings, certificate issues, or server-side blocks.

Common Causes

1. Network Restrictions and Firewalls AI systems often require outbound access to specific server endpoints over HTTPS. If your network blocks these connections or the server is not allowlisted, the AI cannot communicate with the server. Firewalls, content filters, or corporate security policies may prevent access to required ports, such as port 443 for HTTPS traffic, or block specific domains entirely (e.g., ai.portswigger.net for Burp AI) . 2. Proxy or Intercepting Proxy Issues Some networks use upstream proxies or intercepting proxies (like ZScaler) that decrypt and re-sign HTTPS traffic. If the AI system does not trust the proxy's certificate, connections may fail. Configuring the AI to use the correct upstream proxy or importing the proxy certificate into trusted certificates can resolve this . 3. Server-Side Blocks or Access Denied Errors Servers may block requests due to IP restrictions, unusual traffic patterns, or security policies. This can result in errors like “Access Denied” or “403 Forbidden.” Causes include repeated failed login attempts, high request volumes, or regional IP restrictions. These blocks can prevent AI from retrieving data or performing operations . 4. Local Device or Browser Issues Connectivity problems can also stem from the local environment. Browser extensions, URL blockers, VPNs, or antivirus software may interfere with AI access. Cached data or misconfigured browser settings can also cause errors such as “A network error occurred” or “There was a problem preparing your chat” . 5. AI-Specific Configuration Requirements Certain AI platforms, like Azure DevOps MCP Server, require the AI to operate in agent mode and meet system prerequisites (e.g., Node.js 20+). Failure to meet these requirements or misconfiguring the AI assistant can prevent server access .

Troubleshooting Steps

1. Verify general internet connectivity by visiting the server's main website.
2. Check if the AI-specific endpoint is reachable (e.g., ai.portswigger.net).
3. Ensure firewalls, VPNs, and antivirus software are not blocking the connection.
4. Configure upstream proxy settings if your network requires a proxy.
5. Import any necessary certificates for intercepting proxies.
6. Test access from a different browser or device to rule out local issues.
7. Confirm AI platform requirements and agent-mode configuration are met.
8. If the problem persists, collect logs, HAR files, or console errors and contact support for the AI platform . By systematically checking these areas, you can identify whether the issue is network-related, server-side, or due to local configuration, and take appropriate steps to restore AI access.