New Pass-ta-key attack reveals all the things we didn't know about passkeys

The recent buzz surrounding a security vulnerability dubbed Pass-ta-key has sparked a necessary conversation about how we actually store digital identities. A researcher at Palo Alto Networks, Arie Olshtein, outlined a method to extract passkeys from the Google Password Manager app on Windows. This discovery has sent ripples through the security community, challenging the widespread assumption that passkeys are inherently immune to traditional malware attacks.

It is crucial to clarify that this attack is not a fundamental flaw in the passkey standard itself. According to the reporting, the issue lies in implementation rather than the cryptographic protocol. The term Pass-ta-key is a playful blend of passkey and pass-the-hash, but it highlights a specific configuration risk. The confusion stems from a common misconception about where these keys actually live on your device.

Many users and even some security professionals believed that passkeys were stored exclusively in the Trusted Platform Module, or TPM. This hardware enclave is designed to be a hardened, isolated vault for cryptographic keys. The prevailing wisdom was that if a key is in the TPM, malware cannot touch it. This belief provided a false sense of security for those who had migrated away from passwords.

The reality revealed by Olshtein is more nuanced. The Google Password Manager app for Windows stores passkeys in a way that makes them accessible to the application itself. If that application is running on a machine infected with malware, the malicious code can potentially extract the keys. This bypasses the protection of the TPM because the keys are being handled by the software layer before they are ever sent to the hardware vault for signing.

This distinction is vital for understanding the current state of authentication security. Passkeys are not magic bullets that eliminate all risk. They shift the threat model from password theft to device compromise. If an attacker can control the device and the applications running on it, they can intercept the keys during the authentication process. This is not a new problem, but it is a new surface area for existing threats.

The broader implication here is that security is only as strong as its weakest link. Relying solely on the TPM is insufficient if the surrounding software ecosystem is not equally hardened. We need to look at how different password managers and operating systems handle key storage. The industry must move beyond marketing passkeys as a perfect solution and start addressing the implementation details that leave users vulnerable.

What this means for you is that you should audit your password manager settings. If you are using Google Password Manager on Windows, be aware that your passkeys may be more exposed than you think. Consider using a password manager that explicitly stores keys in the TPM or uses a different secure enclave. Additionally, keep your system updated and use robust endpoint protection to prevent malware from accessing your applications.

Here is a prompt you can use with an AI assistant to review your current security posture: Analyze the list of installed password managers and their key storage methods. Identify any that store keys in software rather than hardware enclaves like TPM or Secure Enclave. Provide a risk assessment and suggest alternatives that prioritize hardware-backed storage for critical accounts.