On this page (8 sections)

Security researchers Shubham Shah and Sam Curry reported that the customer login systems behind two car brands' mobile apps could be taken over because an administrator console still accepted its documented default password. According to Infosecurity Magazine, they used that access to take over a test customer account and remotely unlock and start a test car. The two brands were not named, at the automakers' request.

This is not a new product bug. It is a configuration failure that any team running identity software, or any admin panel with a factory password, can check for in minutes. Anyone who operates ForgeRock/PingIDM, an Apache Felix console, or an internal admin page behind a WAF should read on.

What happened

Both brands signed customers in through ForgeRock, an identity platform now part of Ping Identity. The Hacker News reports that the platform runs in two parts, one for login and one, called OpenIDM, that holds customer records. OpenIDM ships with a built-in administrator console powered by Apache Felix. ForgeRock's hardening guide tells operators to remove or protect that console before going live, because it stores its own password separately from the rest of the system.

On these servers the console was still present and still used the documented default: the word admin as both username and password. The researchers published their write-up on November 7. According to The Hacker News, they tested the full chain only on US accounts and on cars they owned or had permission to use, in February 2026.

How the attack worked, at a high level

The console was not open to the internet directly. The reporting says the edge layer in front of the site and the Java server behind it interpreted the same web address differently. That mismatch let a request pass the edge's rules while the server still routed it to the console, where the default password worked. The exact request details are in the researchers' own technical analysis and are deliberately not repeated here.

From the console, the researchers say they ran their own code on the identity server, which gave them reach to the customer records. They changed the email address and password on a test account, signed in through the normal app, and received the same access tokens the real app uses. Those tokens allowed vehicle commands, and the researchers unlocked and started a test car with its owner standing next to it. The session also exposed the car's location and telemetry, per Infosecurity Magazine.

Two details make this harder to notice. Changing the login details did not sign the owner out of their app, and the researchers could divert the app's notifications away from the owner's phone. They wrote that they could have opened a session for any customer account on the affected deployments.

The second chain in Australia

Shah's companion post describes a separate chain against one automaker's Australian systems, researched in July 2026. Per Infosecurity Magazine, it could start from a car's number plate, and combined hardcoded administrator credentials, a WAF bypass and a misconfigured multi-factor authentication step. Exposed data included customers' names, phone numbers and password hashes. The Hacker News adds that the admin credentials had been left in the app's code, intended for a test setup. According to the reports, together the two write-ups covered vehicles in Australia, the US and Canada.

Both outlets say the Australian post does not state when those findings were fixed, and Infosecurity Magazine adds that it does not say when they were reported. Neither write-up says whether anyone else abused the issues. Ping Identity was contacted by Infosecurity Magazine and had not responded at publication.

Who is realistically exposed

  • Teams running ForgeRock OpenIDM or PingIDM, especially older deployments that were never hardened for production.
  • Anyone running an Apache Felix web console, or any framework admin console, on an application server.
  • Organizations that rely on a WAF or reverse proxy path rule as the only thing hiding an admin page.
  • Developers who leave test or administrator credentials in mobile app code or config shipped to customers.

The general lesson reaches beyond cars. Any site with a default login on a management interface, from a CMS to a database panel, is one path-handling quirk away from exposure.

How to check your own exposure

The Hacker News names one concrete check for OpenIDM/PingIDM teams: look for the Apache Felix console at /system/console. The request below prints only the response headers. Run it only against systems you own or are authorized to test, from outside your network and from inside it. A login prompt, redirect to a login page or any console page means the path is reachable and needs action. Replace the hostname with your own.

Check whether the Felix console path responds (run only on your own hosts, read-only)Shell
curl -sI https://your-identity-host.example.com/system/console | head -n 5

Also search your configuration management, container images and mobile app packages for factory passwords and leftover test credentials. Review access logs for requests to the console path, especially unusual encoded or doubled slashes in URLs, and for unexpected changes to customer email addresses or passwords.

What to do now

  1. Inventory every admin console and management interface, including Felix consoles, and note which are reachable from the internet or through a proxy.
  2. Remove the Felix console from production OpenIDM/PingIDM, or at minimum change its default password, as ForgeRock's hardening guidance says.
  3. Do not rely on WAF or edge path rules to hide admin pages. Restrict the console at the application server and network level, for example to a management network or VPN.
  4. Make your edge and application server interpret URLs the same way, and test with encoded and doubled slashes on a staging copy.
  5. Remove hardcoded administrator or test credentials from app code, builds and config, and rotate any that ever shipped.
  6. Review logs for access to the console and for unexplained changes to customer email, password or notification settings.
  7. Invalidate existing sessions and tokens when a credential change is made on an account, so an owner is not left signed in while an attacker takes over.
  8. Check that MFA steps cannot be skipped or misconfigured on account recovery and login flows.
Tip. If you find a default credential in use, treat it as a possible compromise, not just a configuration fix. Rotate secrets, review what the console could reach, and check for changes made through it.

Detection and monitoring afterwards

Alert on any request to management paths from outside your admin network, and on bursts of account email or password changes. For public sites, an outside view helps catch exposed panels: a Webforta Scan of a site you verify can show what is reachable, and certificate, DNS and header-change monitoring can flag unexpected changes. These are general hygiene checks, and they do not replace the server-side fixes above. For background on why edge rules alone are not enough, see what a WAF is and is not.

What we do not know

  • The two automakers' names, which were withheld at their request.
  • Whether anyone else found or used the path before it was closed. Neither write-up reports malicious use.
  • When the Australian findings were reported or fixed.

Sources

This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.

  1. Default Admin Password Let Researchers Unlock and Start Carsinfosecurity-magazine.com ·
  2. Default 'admin:admin' Server Password Helps Researchers Hack Two Automakers, Unlock and Start Carsthehackernews.com ·
  • default credentials
  • ForgeRock
  • OpenIDM
  • WAF bypass
  • account takeover
  • automotive