The Attacker Sits Between the User and the Service

Photo: BleepingComputer

Article

The Attacker Sits Between the User and the Service

ManishankarNovember 22, 20264 min read

Tor's DDoS fight, bank login failures and token-stealing phishing kits all point to the same shift: attacks now target the path to a service, not the service itself.

The common thread in these stories is that the attacks that matter are increasingly aimed at the connective tissue between a user and a service rather than at the service's own code. Tor is defending its network against a sustained DDoS, bank customers are finding their online accounts hard to reach, and phishing kits are being built specifically to capture and replay Microsoft 365 sessions. In each case the attacker is exploiting the dependence of legitimate users on a route, a session or an availability guarantee that sits outside the core product.

Availability Is the Attack Surface

The Tor story is the clearest illustration, as BleepingComputer reported: the project released an emergency update amid an ongoing DDoS attack. A DDoS does not require breaking encryption or stealing credentials. It requires only that enough traffic be directed at a service that legitimate users can no longer get through. For an anonymity network, that is an especially acute pressure point, because the value of the network to its users is inseparable from its being reachable. An emergency update in the middle of an active attack is also a signal about tempo: defenders are patching and hardening while the pressure is still on, rather than after the fact.

When the Front Door Is the Target

The bank story sits on the same axis, though the facts are more limited. As Engadget reported, customers at major banks reported difficulty accessing accounts online, and representatives from the companies described the problem as technical rather than a cyberattack. That attribution matters and should not be discarded. But from a user's perspective the distinction between a technical fault and an attack is often invisible: the account is unreachable either way, and the trust cost is the same. The pattern worth noting is that public-facing access to financial accounts is a single point of failure for a very large number of people, and its failure generates immediate, visible disruption regardless of cause.

Session Theft Moves Past Passwords

The third story explains why the boundary between the user and the service has become so valuable. As BleepingComputer reported, researchers at Huntress broke down phishing kits called EvilTokens, Knight Office and Kali365, which are built to capture and replay Microsoft 365 tokens. Token theft is a meaningful escalation because it does not require the attacker to know or crack a password. If a session token can be captured and replayed, the attacker can present as an already-authenticated user. Multi-factor prompts tied to a fresh login do not necessarily interrupt that, because the session has already been established legitimately.

That is the same underlying logic as the other two stories: the attacker does not need to defeat the service's core defences if it can occupy the channel through which a legitimate user reaches the service. In the Tor case the channel is network availability. In the banking case it is online account access, whatever the cause. In the phishing case it is the authenticated session itself.

Why This Lands Harder in the US

The US market is unusually exposed to this pattern for structural reasons visible in the material. Microsoft 365 is the collaboration backbone for a very large share of American enterprises, so a kit that reliably captures and replays its sessions has a broad addressable target set without the attackers needing to customise per victim. Consumer banking in the US is heavily digital-first, so online access difficulty is not a minor inconvenience but a direct block on routine financial activity for customers of major institutions. And US technology companies that operate public-facing platforms carry the availability burden directly: they absorb the cost of mitigation, the reputational hit of downtime and the engineering distraction of emergency updates, as Tor's situation shows.

For US consumers, the practical consequence is that familiar signals of safety are becoming less reliable. A padlock, a working login page and a successful multi-factor prompt no longer establish that a session is clean. For US technology companies, the consequence is that investment in availability and in session integrity has to be treated as core security work rather than as operational overhead.

Detection Favours the Unglamorous

The Huntress work, as reported by BleepingComputer, is notable for including guidance on how defenders can detect and contain these kits rather than only describing the tooling. That emphasis is the right one. If tokens can be replayed, then detection has to focus on session behaviour: where sessions originate, how they move and whether a session's activity is consistent with the user behind it. Containment likewise has to include the ability to invalidate sessions, not just reset passwords.

On the availability side, the Tor example points to a different but related discipline: the willingness to ship an emergency update while an attack is in progress. That is an operational posture, and it is expensive to maintain. The bank story, with companies attributing the disruption to a technical problem, is a reminder that clear public communication about cause is part of the same problem, because users cannot distinguish an outage from an attack on their own.

What to Watch

The material points to a few concrete things to track. First, whether Tor's emergency update measurably changes the course of the ongoing DDoS, which would say something about how much mitigation can be achieved mid-attack. Second, how the banks' technical problem resolves and whether the companies provide more detail on cause, given that they have already stated it was not a cyberattack. Third, whether detection and containment guidance for token-replay kits such as EvilTokens, Knight Office and Kali365 gets adopted into routine practice at organisations running Microsoft 365, particularly around session invalidation rather than password resets alone.

The larger question is whether US companies treat the path to the service as part of the service. The evidence in these three stories suggests attackers already do.

Sources: BleepingComputer, Engadget.

More on this beat: Cybersecurity on TechManNews.

#cyber attacks#DDoS#phishing#session tokens#account access#US enterprise security

Newsletter

Get Tech News in Your Inbox

The latest AI, gadgets, software and startup stories from TechManNews, delivered every morning - free.