What are device-bound session credentials, and how do they stop the newest form of account takeover?

Google's Chrome browser has begun rolling out a new security mechanism called device-bound session credentials, a defense researchers and security teams describe as among the strongest protections yet against a rapidly growing category of account takeover attack. The feature targets a specific, increasingly common technique: stealing the session tokens that keep a user logged in, rather than stealing their password.
To understand why that distinction matters, it helps to know how most modern logins actually work behind the scenes. After you enter a password and complete any additional verification step, your browser doesn't ask you to re-authenticate every time you click a link — instead, it stores a session token, a small piece of data that proves to the website you've already logged in successfully. That token typically lives in the browser as a cookie, quietly authenticating every subsequent request for as long as the session stays active.
That convenience has become a major security liability. As more services adopted multi-factor authentication over the past several years, making stolen passwords alone far less useful to attackers, cybercriminals increasingly shifted their focus to stealing session tokens directly, using malware built specifically to harvest them from an infected computer. A stolen session token lets an attacker walk straight past the login screen entirely — no password, no verification code required, because as far as the website is concerned, the session was already authenticated.
Device-bound session credentials are designed to close exactly that gap. Rather than treating a session token as a portable secret that works from anywhere it's copied to, the new mechanism cryptographically ties the session to a private key generated and stored on the specific device where the user logged in — typically inside secure hardware built into the device that's designed to make the key extremely difficult to extract. Each time the browser needs to prove the session is still valid, it uses that device-bound key to sign the request.
That single design choice changes the economics of the attack. If a piece of malware manages to steal a session cookie or token off a victim's machine and an attacker tries to use it from their own computer elsewhere, the website's server can check whether the request carries a valid signature from the original device's key — and reject it if it doesn't. The stolen token, on its own, becomes essentially worthless outside the device it was generated on, undermining the entire premise of the theft.
Chrome's implementation builds the mechanism directly into the browser, giving website operators a relatively straightforward path to adopt the protection without having to build the underlying cryptographic infrastructure themselves. Sites that support the feature can begin issuing device-bound sessions to Chrome users, extending the protection across whichever services choose to implement it on their end.
The scale of the problem this is meant to address is considerable. Infostealer malware — malicious software built specifically to harvest saved passwords, browser cookies and session tokens from an infected machine — has become a significant part of the cybercrime economy, with stolen session data bought and sold in bulk on criminal marketplaces. Phishing kits, too, have increasingly shifted toward capturing live sessions in real time rather than just credentials, precisely because a captured session can skip past multi-factor authentication entirely.
Security researchers frame device-bound session credentials as closing a gap that earlier generations of defenses didn't fully address. Passwords alone were vulnerable to phishing and reuse; multi-factor authentication meaningfully raised the bar but didn't stop an attacker who could steal an already-authenticated session; tying that session to a specific piece of hardware removes the portability that made stolen sessions valuable to begin with.
The protection does have limits worth understanding. It's designed to stop a stolen session token from being replayed on a different device — it doesn't protect against malware that remains active and operating directly on the legitimate device itself, where an attacker with that level of access could still, in principle, act through the authenticated session locally rather than trying to move it elsewhere. It's a defense against exfiltration and remote reuse, not a cure for every form of device compromise.
Even with that caveat, security researchers describe the shift as a meaningful step forward in browser security, one that directly targets a technique that has become one of the most effective tools in an attacker's arsenal. As Chrome's implementation matures and more websites adopt the underlying standard, the approach is expected to spread to other major browsers as well, part of a broader industry trend toward anchoring authentication in hardware-backed proof rather than secrets that can simply be copied and carried away.
Read next

Made by Google 2026: what to expect from today's Pixel 11 launch event
Google is set to unveil a wave of new Pixel devices at its annual Made by Google event on August 12th. Pre-event leaks have offered an unusually detailed preview of what's expected — from a broader Pixel 11 phone lineup to new wearables and possibly a whole new accessory category for the company.

How new surveillance tech links your phone to your license plate
Researchers report that roadside cameras photographing license plates are increasingly paired with equipment that also captures wireless signals from nearby phones. That combination can turn a simple 'this car passed here' record into a much richer profile of who was actually inside.

Gemini hits 1 billion users: how people are actually using Google's AI chatbot
Google announced that its flagship consumer AI app, Gemini, has surpassed 1 billion users. The company also revealed that 63% of Gemini users talk directly to the assistant using its voice feature, and that Gemini now generates more than 150 million images every day.

What is the Riemann hypothesis, and how close is AI to solving it?
On the Riemann hypothesis, one of mathematics' most famous unsolved problems for more than 150 years, an unreleased Anthropic AI model reportedly made significant progress. The model didn't prove the hypothesis, but went further than many experts expected an AI system to get.

What are small language models, and why are they starting to rival giant AI systems?
Liquid AI's new LFM2.5 model, with just 2.6 billion parameters, is competitive with models four times its size. It's the latest example of a fast-growing small language model movement challenging the assumption that bigger is always better in AI.