Most phishing websites are little more than static copies of login pages for popular online destinations, and they are often quickly taken down by anti-abuse activists and security firms. But a stealthy new phishing-as-a-service offering lets customers sidestep both of these pitfalls: It uses cleverly disguised links to load the target brand’s real website, and then acts as a relay between the target and the legitimate site — forwarding the victim’s username, password and multi-factor authentication (MFA) code to the legitimate site and returning its responses.

There are countless phishing kits that would-be scammers can use to get started, but successfully wielding them requires some modicum of skill in configuring servers, domain names, certificates, proxy services, and other repetitive tech drudgery. Enter Starkiller, a new phishing service that dynamically loads a live copy of the target login page and records everything the user types, proxying the data to the legitimate site and back to the victim.

According to an analysis of Starkiller by the security firm Abnormal AI, the service lets customers select a brand to impersonate (e.g., Apple, Facebook, Google, Microsoft et. al.) and generates a deceptive URL that visually mimics the legitimate domain while routing traffic through the attacker’s infrastructure.

For example, a phishing link targeting Microsoft customers appears as “login.microsoft.com@[malicious/shortened URL here].” The “@” sign in the link trick is an oldie but goodie, because everything before the “@” in a URL is considered username data, and the real landing page is what comes after the “@” sign. Here’s what it looks like in the target’s browser:

Once Starkiller customers select the URL to be phished, the service spins up a Docker container running a headless Chrome browser instance that loads the real login page, Abnormal found.

“The container then acts as a man-in-the-middle reverse proxy, forwarding the end user’s inputs to the legitimate site and returning the site’s responses,” Abnormal researchers Callie Baron and Piotr Wojtyla wrote in a blog post on Thursday. “Every keystroke, form submission, and session token passes through attacker-controlled infrastructure and is logged along the way.”

Starkiller in effect offers cybercriminals real-time session monitoring, allowing them to live-stream the target’s screen as they interact with the phishing page, the researchers said.

“The platform also includes keylogger capture for every keystroke, cookie and session token theft for direct account takeover, geo-tracking of targets, and automated Telegram alerts when new credentials come in,” they wrote. “Campaign analytics round out the operator experience with visit counts, conversion rates, and performance graphs—the same kind of metrics dashboard a legitimate SaaS [software-as-a-service] platform would offer.”

Abnormal said the service also deftly intercepts and relays the victim’s MFA credentials, since the recipient who clicks the link is actually authenticating with the real site through a proxy, and any authentication tokens submitted are then forwarded to the legitimate service in real time.

“The attacker captures the resulting session cookies and tokens, giving them authenticated access to the account,” the researchers wrote. “When attackers relay the entire authentication flow in real time, MFA protections can be effectively neutralized despite functioning exactly as designed.”

The “URL Masker” feature of the Starkiller phishing service features options for configuring the malicious link. Image: Abnormal.

Starkiller is just one of several cybercrime services offered by a threat group calling itself Jinkusu, which maintains an active user forum where customers can discuss techniques, request features and troubleshoot deployments. One a-la-carte feature will harvest email addresses and contact information from compromised sessions, and advises the data can be used to build target lists for follow-on phishing campaigns.

This service strikes me as a remarkable evolution in phishing, and its apparent success is likely to be copied by other enterprising cybercriminals (assuming the service performs as well as it claims). After all, phishing users this way avoids the upfront costs and constant hassles associated with juggling multiple phishing domains, and it throws a wrench in traditional phishing detection methods like domain blocklisting and static page analysis.

It also massively lowers the barrier to entry for novice cybercriminals, Abnormal researchers observed.

“Starkiller represents a significant escalation in phishing infrastructure, reflecting a broader trend toward commoditized, enterprise-style cybercrime tooling,” their report concludes. “Combined with URL masking, session hijacking, and MFA bypass, it gives low-skill cybercriminals access to attack capabilities that were previously out of reach.”

For the past week, the massive “Internet of Things” (IoT) botnet known as Kimwolf has been disrupting The Invisible Internet Project (I2P), a decentralized, encrypted communications network designed to anonymize and secure online communications. I2P users started reporting disruptions in the network around the same time the Kimwolf botmasters began relying on it to evade takedown attempts against the botnet’s control servers.

Kimwolf is a botnet that surfaced in late 2025 and quickly infected millions of systems, turning poorly secured IoT devices like TV streaming boxes, digital picture frames and routers into relays for malicious traffic and abnormally large distributed denial-of-service (DDoS) attacks.

I2P is a decentralized, privacy-focused network that allows people to communicate and share information anonymously.

“It works by routing data through multiple encrypted layers across volunteer-operated nodes, hiding both the sender’s and receiver’s locations,” the I2P website explains. “The result is a secure, censorship-resistant network designed for private websites, messaging, and data sharing.”

On February 3, I2P users began complaining on the organization’s GitHub page about tens of thousands of routers suddenly overwhelming the network, preventing existing users from communicating with legitimate nodes. Users reported a rapidly increasing number of new routers joining the network that were unable to transmit data, and that the mass influx of new systems had overwhelmed the network to the point where users could no longer connect.

When one I2P user asked whether the network was under attack, another user replied, “Looks like it. My physical router freezes when the number of connections exceeds 60,000.”

A graph shared by I2P developers showing a marked drop in successful connections on the I2P network around the time the Kimwolf botnet started trying to use the network for fallback communications.

The same day that I2P users began noticing the outages, the individuals in control of Kimwolf posted to their Discord channel that they had accidentally disrupted I2P after attempting to join 700,000 Kimwolf-infected bots as nodes on the network.

The Kimwolf botmaster openly discusses what they are doing with the botnet in a Discord channel with my name on it.

Although Kimwolf is known as a potent weapon for launching DDoS attacks, the outages caused this week by some portion of the botnet attempting to join I2P are what’s known as a “Sybil attack,” a threat in peer-to-peer networks where a single entity can disrupt the system by creating, controlling, and operating a large number of fake, pseudonymous identities.

Indeed, the number of Kimwolf-infected routers that tried to join I2P this past week was many times the network’s normal size. I2P’s Wikipedia page says the network consists of roughly 55,000 computers distributed throughout the world, with each participant acting as both a router (to relay traffic) and a client.

However, Lance James, founder of the New York City based cybersecurity consultancy Unit 221B and the original founder of I2P, told KrebsOnSecurity the entire I2P network now consists of between 15,000 and 20,000 devices on any given day.

An I2P user posted this graph on Feb. 10, showing tens of thousands of routers — mostly from the United States — suddenly attempting to join the network.

Benjamin Brundage is founder of Synthient, a startup that tracks proxy services and was the first to document Kimwolf’s unique spreading techniques. Brundage said the Kimwolf operator(s) have been trying to build a command and control network that can’t easily be taken down by security companies and network operators that are working together to combat the spread of the botnet.

Brundage said the people in control of Kimwolf have been experimenting with using I2P and a similar anonymity network — Tor — as a backup command and control network, although there have been no reports of widespread disruptions in the Tor network recently.

“I don’t think their goal is to take I2P down,” he said. “It’s more they’re looking for an alternative to keep the botnet stable in the face of takedown attempts.”

The Kimwolf botnet created challenges for Cloudflare late last year when it began instructing millions of infected devices to use Cloudflare’s domain name system (DNS) settings, causing control domains associated with Kimwolf to repeatedly usurp Amazon, Apple, Google and Microsoft in Cloudflare’s public ranking of the most frequently requested websites.

James said the I2P network is still operating at about half of its normal capacity, and that a new release is rolling out which should bring some stability improvements over the next week for users.

Meanwhile, Brundage said the good news is Kimwolf’s overlords appear to have quite recently alienated some of their more competent developers and operators, leading to a rookie mistake this past week that caused the botnet’s overall numbers to drop by more than 600,000 infected systems.

“It seems like they’re just testing stuff, like running experiments in production,” he said. “But the botnet’s numbers are dropping significantly now, and they don’t seem to know what they’re doing.”

Microsoft today released updates to fix more than 50 security holes in its Windows operating systems and other software, including patches for a whopping six “zero-day” vulnerabilities that attackers are already exploiting in the wild.

Zero-day #1 this month is CVE-2026-21510, a security feature bypass vulnerability in Windows Shell wherein a single click on a malicious link can quietly bypass Windows protections and run attacker-controlled content without warning or consent dialogs. CVE-2026-21510 affects all currently supported versions of Windows.

The zero-day flaw CVE-2026-21513 is a security bypass bug targeting MSHTML, the proprietary engine of the default Web browser in Windows. CVE-2026-21514 is a related security feature bypass in Microsoft Word.

The zero-day CVE-2026-21533 allows local attackers to elevate their user privileges to “SYSTEM” level access in Windows Remote Desktop Services. CVE-2026-21519 is a zero-day elevation of privilege flaw in the Desktop Window Manager (DWM), a key component of Windows that organizes windows on a user’s screen. Microsoft fixed a different zero-day in DWM just last month.

The sixth zero-day is CVE-2026-21525, a potentially disruptive denial-of-service vulnerability in the Windows Remote Access Connection Manager, the service responsible for maintaining VPN connections to corporate networks.

Chris Goettl at Ivanti reminds us Microsoft has issued several out-of-band security updates since January’s Patch Tuesday. On January 17, Microsoft pushed a fix that resolved a credential prompt failure when attempting remote desktop or remote application connections. On January 26, Microsoft patched a zero-day security feature bypass vulnerability (CVE-2026-21509) in Microsoft Office.

Kev Breen at Immersive notes that this month’s Patch Tuesday includes several fixes for remote code execution vulnerabilities affecting GitHub Copilot and multiple integrated development environments (IDEs), including VS Code, Visual Studio, and JetBrains products. The relevant CVEs are CVE-2026-21516, CVE-2026-21523, and CVE-2026-21256.

Breen said the AI vulnerabilities Microsoft patched this month stem from a command injection flaw that can be triggered through prompt injection, or tricking the AI agent into doing something it shouldn’t — like executing malicious code or commands.

“Developers are high-value targets for threat actors, as they often have access to sensitive data such as API keys and secrets that function as keys to critical infrastructure, including privileged AWS or Azure API keys,” Breen said. “When organizations enable developers and automation pipelines to use LLMs and agentic AI, a malicious prompt can have significant impact. This does not mean organizations should stop using AI. It does mean developers should understand the risks, teams should clearly identify which systems and workflows have access to AI agents, and least-privilege principles should be applied to limit the blast radius if developer secrets are compromised.”

The SANS Internet Storm Center has a clickable breakdown of each individual fix this month from Microsoft, indexed by severity and CVSS score. Enterprise Windows admins involved in testing patches before rolling them out should keep an eye on askwoody.com, which often has the skinny on wonky updates. Please don’t neglect to back up your data if it has been a while since you’ve done that, and feel free to sound off in the comments if you experience problems installing any of these fixes.