LockBit 5.0, Chrome zero-day and infostealers expose identity gaps

At a glance

What this is: Anomali’s Cyber Watch rounds up active LockBit 5.0 ransomware, Chrome zero-day exploitation, and OpenClaw-targeting infostealers as evidence that exposed credentials and weak trust boundaries remain a persistent attack path.

Why it matters: It matters because IAM, NHI, and endpoint teams are all implicated when attackers can move from a browser flaw or commodity malware into tokens, keys, and remote access pathways.

Context

This roundup sits at the intersection of endpoint security, browser risk, and identity governance. The common problem is not a single exploit class but the persistence of credentials, tokens, and trusted context across systems that users and security teams often treat separately.

For IAM and NHI practitioners, the relevance is direct. Ransomware still leverages access paths, browser exploitation still creates account and session exposure, and AI-adjacent local assistants are now storing authentication material and user context in places commodity malware can reach.

Key questions

Q: What breaks when browser exploits reach identity sessions and cloud consoles?

A: A browser exploit can do more than run code in a sandbox. If the endpoint already holds SSO sessions, saved credentials, or cloud admin tabs, the attacker may inherit trusted access without needing to defeat MFA again. That shifts the problem from patching alone to session hygiene, token protection, and rapid revocation of active access artefacts.

Q: Why do infostealers pose a special risk to local AI assistants?

A: Infostealers can harvest configuration files, gateway tokens, device keys, and cached context without understanding the application. Local AI assistants are risky because those artifacts often sit together on disk and can be enough to impersonate a user or reconnect to a trusted service. The problem is durable trust material, not the malware family itself.

Q: What signs show that ransomware may be targeting virtual infrastructure?

A: Warning signs include campaigns that target Windows, Linux, and ESXi in the same operation, log tampering, and activity that reaches hypervisor layers rather than only endpoints. When ransomware can affect a virtualisation host, one compromise can cascade into multiple workloads, so defenders should watch for access paths into the management plane.

Q: How should teams balance browser patching with IAM controls?

A: Both matter, but the right priority is to reduce what an attacker can do with a stolen session. Browser patching closes the exploit path, while IAM controls limit the value of any session or token that is already exposed. Teams should align endpoint, IAM, and SOC response so each sees the same trust boundary.

Technical breakdown

Why browser exploitation becomes an identity problem

A browser zero-day is not just a code execution issue when the browser is the front door to federated sessions, SaaS consoles, and internal portals. Once an attacker can run code in the renderer process, they may be able to ride the victim’s authenticated context, access session-bearing sites, or chain into phishing and token theft. The key governance failure is assuming browser compromise stops at the browser boundary. In practice, the browser is an identity broker for modern work, which means a successful exploit can become an authentication event, not just an endpoint event.

Practical implication: treat browser exploitation as potential session compromise and verify that identity and endpoint teams share containment playbooks.

How infostealers turn local AI assistants into trust targets

Commodity infostealers do not need bespoke logic to create risk. They can sweep directories, harvest configuration files, and collect tokens, device keys, and cached context that were never meant to sit together on disk. In an AI assistant or agentic workflow, that material can be enough to impersonate a user, reconnect to a gateway, or reuse paired-device trust. The architectural issue is that local convenience often outruns lifecycle control: the secrets, keys, and memory artifacts are durable even when the session is meant to be ephemeral.

Practical implication: classify AI assistant config files, gateway tokens, and device keys as sensitive identity material, not harmless application data.

How ransomware campaigns exploit overexposed machine access

LockBit 5.0 shows that ransomware operators still value broad platform reach, stealth, and disruption over simple file encryption. When attackers can hit Windows, Linux, and ESXi from the same campaign, the real target is operational dependency, not just data at rest. Virtualisation layers matter because a compromised host can affect multiple workloads at once, and double-extortion increases the leverage of stolen data. The technical lesson is that availability failures often begin as identity failures, because the initial foothold and later movement depend on access that was broader than it needed to be.

Practical implication: map privileged access paths into virtual infrastructure and reduce standing rights before ransomware operators can convert them into blast radius.

Threat narrative

Attacker objective: The attacker seeks durable access to trusted identities and operational systems, then uses that access to steal data, spread control, or extort victims.

1. Entry occurs through a malicious browser page, an infostealer run on an endpoint, or initial ransomware access into exposed infrastructure.
2. Credential_harvested or trust data is then collected from browser sessions, OpenClaw configuration files, or machine and administrative environments.
3. Escalation follows when stolen tokens, keys, or privileged access let the attacker impersonate users, reach connected services, or deploy ransomware across multiple platforms.
4. Impact is reached when the attacker steals data, encrypts systems, or uses the compromised trust material to widen the breach surface.

Breaches seen in the wild

• ClawHub malicious skills 2026: Hundreds of malicious ClawHub skills for the OpenClaw agent pushed credential stealers in early 2026; Koi found 341 of 2,857 in the ClawHavoc campaign.
• Cisco Active Directory credentials leak 2025: Kraken leaked Cisco Active Directory hashes, including service and krbtgt accounts; Cisco says they came from its 2022 breach, not a new one.

Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.

NHI Mgmt Group analysis

Identity data now behaves like shared infrastructure: browser sessions, device keys, gateway tokens, and admin credentials are increasingly reusable across otherwise separate attack paths. That is why a browser zero-day, a commodity infostealer, and ransomware activity belong in the same governance conversation. The control failure is not just exposure, it is the persistence of trust material after the original task is done, and practitioners should govern it as a lifecycle problem.

AI-adjacent workflows are creating a new trust gap: locally run assistants and related tools can accumulate identity artifacts that attackers do not need to understand deeply before they can exploit them. A file-grabber infostealer can still capture enough material to impersonate a user or reach a linked service, which means the security model must account for configuration, tokens, and cached context as governed assets. This is where NHI governance meets agentic workflow risk, and the boundary is already operational.

Trusted context is the new blast radius: the article’s common thread is not malware sophistication alone but the amount of damage available once an attacker reaches an authenticated environment. That changes the question from whether the initial exploit was complex to whether the environment allowed that exploit to inherit too much authority. NHI and IAM teams should interpret this as a reason to shrink the privileges attached to sessions, tokens, and machine-bound trust.

Detection has to move earlier in the trust chain: by the time ransomware or infostealers are active, the underlying weakness is usually long before response begins. Browser restart compliance, token exposure, and over-permissive service access are governance signals, not just operational details. Practitioners should treat the article as evidence that identity controls must be enforced at issuance, storage, and session boundary, not only at incident response time.

What this signals

Chrome zero-days now sit inside identity governance scope: the practical question is no longer whether the browser is patched, but whether a compromised browser can inherit enough trust to reach high-value services. That makes session design, device posture, and restart enforcement part of identity risk management.

OpenClaw-style incidents show why AI workflow artefacts need lifecycle control: tokens, keys, and cached context are now operational secrets, even when they are produced by local tools rather than central identity platforms. Practitioners should assume any file that can reconnect an assistant to a service can also reconnect an attacker.

Trusted context is becoming the limiting resource: attackers do not need to own every layer if they can reuse one authenticated layer long enough to move laterally. That means NHI and IAM programmes should measure how much authority sits inside a session, not just how many credentials exist.

For practitioners

• Harden browser restart compliance Verify that patched Chrome builds are not just downloaded but actually restarted so the fix takes effect across managed endpoints.
• Classify AI assistant secrets as governed assets Inventory local configuration files, gateway tokens, device keys, and memory artefacts created by AI assistants and move them under the same handling rules as other sensitive credentials.
• Reduce standing privilege into virtual infrastructure Review administrator and service access to ESXi and other virtualisation layers so ransomware crews cannot turn a single foothold into broad workload impact.
• Monitor for browser and infostealer indicators Alert on post-exploitation behaviours such as token theft, unusual session reuse, and malware accessing directories associated with local AI workflows or credential stores.

Key takeaways

• The common risk across ransomware, browser exploitation, and infostealers is the same: attackers keep turning trusted context into a reusable access path.
• The article shows that browser sessions, local AI configuration files, and virtual infrastructure can all become identity exposure points, even when the initial threat looks different.
• Teams should tighten session governance, reduce standing privilege, and treat AI assistant secrets as governed credentials rather than ordinary application files.

Key terms

• Browser Session Theft: Browser session theft is the capture of authenticated web-session state after a user has already signed in. It is dangerous because the attacker may not need to know the password or defeat MFA again, only reuse the live session to act as the victim.
• Infostealer: An infostealer is malware built to collect credentials, session material, tokens, and other authentication data from infected systems. In NHI programmes, the risk is not only theft but reuse, because harvested workload secrets can unlock cloud access long after the initial infection.
• Machine Identity: The digital identity of a machine, device, or workload, such as a server, container, or VM, used to authenticate it within a network. Sometimes used interchangeably with NHI, though NHI is the broader category.
• Standing Privilege: Standing privilege is access that remains active even when no immediate task requires it. For NHI programmes, it is a common failure mode because long-lived credentials and persistent roles create unnecessary exposure. Reducing standing privilege usually means tighter expiry, on-demand access, and clearer review of who or what still needs access.

Deepen your knowledge

NHI Foundation Level course, the industry's only accredited NHI security programme, covers NHI governance, machine identity security, and secrets management. It gives practitioners a common control model for the identity risks that now span browsers, workloads, and AI-adjacent tools.