A phony "better-than-Binance job offer" cost Binance $6.3 billion.

Compiled & Written by: KarenZ, Foresight News

Back in June, a Binance wallet engineer searching for AI coding tools on a search engine downloaded a tampered software package. Months later, a second engineer joined an online job interview and downloaded a code assignment as requested by the interviewer. A month on from that, attackers had gained the ability to take control of both employees’ devices via the firm’s internal AI tools, and obtained permission credentials that granted access to core infrastructure.

This is not the plot of a cybersecurity thriller, but the reconstructed attack path laid out in Binance’s initial investigation report on the October 29 security incident, which the firm released on November 3.

The incident involved roughly $6.34 billion in digital assets. Binance noted that attackers leveraged unauthorized access to the firm’s internal wallet signing service to initiate transactions outside the standard withdrawal workflow. The U.S. Federal Bureau of Investigation (FBI) has attributed the incident to North Korea.

But fixating only on the final asset transfer risks missing the most alarming dimension of this attack: the attackers did not break into Binance abruptly on October 29. Instead, they began hunting for entry points months in advance, gradually compromising employee devices, gaining internal access capabilities, and leveraging internal corporate tools to move closer to the firm’s wallet infrastructure.

Binance has framed this process as a Multi-stage Campaign, explicitly stating that it was neither an isolated system breach nor a simple phishing attack.

Starting with the supply chain attack in June, moving to the fake recruitment scam in September, and ending with leveraging internal AI tools to expand access scope in October, the attackers pulled off a combined penetration targeting personnel, devices, software, and permission systems.

This attack chain has also sounded an alarm bell for the entire crypto industry: when an exchange’s core asset security is tied to the development tools, AI applications, and internal permission systems used daily by its employees, attackers do not necessarily need to breach the firm’s sturdiest perimeter first.

Sometimes, all it takes is tricking one employee into downloading an apparently legitimate software package.

June: The Incident Begins With a Tampered Homebrew Installer Package

The attack preparation timeline disclosed by Binance dates back to June 2026 at the earliest.

At that time, an engineer on the wallet DevOps team (referred to as Engineer A in Binance’s report) was searching for a popular AI coding tool via a search engine, and downloaded a tampered Homebrew installer package.

Homebrew is a widely used package manager that allows developers to quickly install a range of development software and tools.

But if an installer package is maliciously tampered with, the otherwise routine software installation process can become a channel for malicious code to infiltrate devices. This type of attack is known as a Supply Chain Attack.

Unlike traditional phishing attacks, supply chain attacks do not necessarily rely on directly tricking users into revealing their passwords. Instead, they embed malicious code into software, tools, or update processes that users trust, causing users to execute malicious programs during their normal software usage.

This risk is particularly salient for developers, whose work devices typically host a large number of third-party tools, and may store access credentials for code repositories, cloud services, and internal infrastructure.

Binance noted that its security team detected anomalies at the time, proactively replaced Engineer A’s device, and rotated all relevant credentials.

However, the attackers had already gained visibility into the engineer’s job role, as well as information about his relevant colleagues on the wallet infrastructure team.

In other words, even with the device replaced and credentials rotated, the attackers could continue searching for other entry points to attempt access to the same team.

September: A Fake Recruitment Scam Delivered Malicious Code to an Employee’s Work Device

In September, the attackers shifted their target to another software engineer (referred to as Engineer B in the report), deploying a more deceptive social engineering attack.

Engineer B had publicly posted his professional experience on a legitimate job platform, including his tenure at Binance. Shortly after, an individual claiming to be a Project Director at an unnamed company reached out to him directly to invite him for an interview.

The entire recruitment process was crafted to appear highly legitimate and complete.

The first round of the interview was conducted by an individual claiming to be an HR representative, with both parties communicating via real-time video call. To build trust, the interviewer shared a social media account containing personal photos and videos.

After the first round concluded, the attackers moved the process forward.

The second round of the interview was moved to a popular social chat application. As the interview drew to a close, the interviewer asked Engineer B to download a code repository labeled "Assignment" and run it on his work device.

For software engineers, downloading code, setting up development environments, and completing technical assessments are standard parts of a routine recruitment process.

In reality, this code repository contained malicious code. Per Binance’s disclosure, the attackers used code obfuscation and detection evasion techniques, including replacing components of legitimate development tools with malicious versions, to enable command execution and establish communication with external servers controlled by the attackers.

Binance described the tool as 0-day malware, which was not detected by endpoint security protection tools at the time. The attackers also attempted to hide malicious activity and erase related traces to increase the difficulty of subsequent detection and investigation.

• Binance even publicly disclosed the specific operations the malicious script performed on the pyenv tool installed on Engineer B’s device:
• Create a new pyenv_ file;
• Replace the original legitimate pyenv component with a malicious version;
• Delete the original legitimate file;
• Rename the malicious file back to pyenv;
• Delete the original command entry, recreate the symbolic link, and launch the zsh shell.

pyenv is a common Python version management tool. By replacing its legitimate components, the attackers disguised malicious programs as normal development tools, enabling them to execute commands and connect to external servers.

The covert nature of this type of attack is that users may still be able to use their familiar development commands as normal, without knowing that the underlying program has been replaced.

A seemingly routine technical interview thus became the entry point for attackers to gain access to an employee’s work device.

October 20: Internal AI Tools Become Entangled in the Attack Chain

If the first two phases focused on gaining initial access to employee devices, the events of October 20 meant the attackers began expanding their reach deeper into Binance’s wallet team.

Binance disclosed that the attackers leveraged Engineer B’s already compromised device, as well as an internal Binance AI tool, to further take control of the devices belonging to Engineer A and another wallet software engineer, Engineer C.

Binance did not disclose the specific name of this AI tool, nor did it specify exactly which function the attackers exploited. As such, it is not possible to conclude that the AI model itself has vulnerabilities, nor to reduce this incident to a simple prompt injection attack.

What is confirmed is that the internal AI tool was incorporated into the attackers’ path to expand their access scope.

As AI tools are adopted for internal corporate use, they may connect to code repositories, knowledge bases, development environments, and other internal services. The access permissions, identity authentication mechanisms, and operation logs held by these tools thus become part of the enterprise security framework.

This incident also exposed another critical issue.

Binance noted that the attackers soon discovered a Container Deployment Permission Token stored on Engineer A’s device.

A Container is a relatively isolated software runtime environment where development teams can package applications and their dependencies for deployment to servers or cloud platforms.

The Container Deployment Permission Token can be understood as a passkey to invoke relevant deployment capabilities. It is not a wallet private key, but if it grants access to critical internal environments, it can serve as an entry point to move closer to core systems.

By this point, the attackers had not only taken control of employee devices, but also obtained access credentials that could connect to internal deployment systems.

October 29: From Employee Devices to the Wallet Signing Service

On October 29, the attack entered its final phase.

Per Binance’s disclosure, between 17:15 and 19:56 UTC that day, the attackers logged into Engineer C’s device, deployed a container using the credentials previously stolen from Engineer A, triggered the malicious script, and formally launched the attack.

After that, the attackers leveraged unauthorized access to Binance’s internal wallet signing service to initiate transactions outside the standard withdrawal workflow.

Binance stated that the incident involved roughly $6.34 billion in digital assets. The attackers also leveraged Ethereum’s EIP-7702 authorization mechanism to grant their controlled smart contracts partial execution permissions over Binance’s hot wallets.

57 Suspected North Korean Attack Attempts: The Crypto Industry Must Reassess Its Internal Security Boundaries

Binance also disclosed in its report that, since 2026, its security team has detected and blocked 57 cyber threat incidents suspected of being linked to North Korea.

These attack attempts involved tactics including fake job interviews, malware disguised as popular development tools, tampered software packages and browser extensions, fake video conferencing applications, as well as credential theft via social engineering or remote access malware deployment. A prior public announcement from the FBI also attributed the Binance incident to North Korea, labeling the related malicious cyber activity TraderTraitor.

This means that attacks targeting crypto companies are no longer limited to hunting for smart contract vulnerabilities or stealing user private keys. Attackers are also continuously studying corporate internal personnel, software environments, and permission systems.

For exchanges, traditional security perimeters need to be extended further into internal networks.

Employee devices require stricter isolation and monitoring; development tools and software packages need source verification; internal AI tools should follow the principle of least privilege; and credentials such as container deployment tokens should not be stored long-term on regular employee devices.

More importantly, even if employee accounts and devices are compromised, attackers should not be able to easily move laterally from one work environment to the wallet signing system.

This incident is not a simple story of inherent vulnerabilities in EIP-7702 or AI tools, but a combined attack centered on personnel, devices, software, and permissions.

In the past, the crypto industry’s biggest concern was stolen private keys. Today, a more alarming question emerges: even if private keys are not leaked, can attackers, by controlling internal processes, trick the system into performing operations it should not execute?

This is perhaps the most critical security lesson the Binance incident leaves for the entire industry.

Regarding follow-up measures and rectifications, Binance stated that it has hired independent third-party cybersecurity experts to conduct a comprehensive review of the incident, and will increase investment in AI security detection, endpoint protection, developer environment security, and anomalous behavior monitoring. At the same time, Binance is establishing a security research fund, and collaborating with leading AI firms, industry participants, and security companies to launch a dedicated program focused on AI security in blockchain environments.