News linked to both this project and an event.
Odaily reports: Cursor, the AI coding tool under SpaceX, posted on X: "We have discovered a security incident affecting a tampered Claude Code VS Code extension accessed through our Cursor IDE extension marketplace, affecting up to 5,000 users. This malicious extension version was not developed by Anthropic, and it does not affect other extensions downloaded from the official website, Grok Bot, Grok, or our other SpaceXSI products.In response, we shut down the extension marketplace this afternoon, worked with the relevant service providers to remove the malware, and reopened the marketplace after confirming there was no further risk. If you are affected, please refer to the instructions in the official blog post to learn how to identify the malicious package and take measures to mitigate any potential impact. We have asked all customers to update Claude Code for VS Code to version 2.1.314."
according to Bitcoin News monitoring, attackers allegedly used EIP-7702 delegation to point the Binance Wallet to malicious code, thereby transferring funds. How the attacker obtained authorization is still under investigation, with two possibilities currently: the private key was stolen, or the attacker obtained a malicious authorization signature through a compromised signing process. Neither scenario has been confirmed. If the private key was not stolen, this attack may expose a significant vulnerability in the exchange's transaction authorization mechanism, rather than a private key security issue.
Cybersecurity firm XBOW disclosed that it discovered a critical remote code execution (RCE) vulnerability in a publicly accessible web application for the U.S. Election Assistance Commission (EAC), which can be exploited without authentication. The application is used for resource and logistics calculations, and is not a voting machine or voter registration system.
Researchers disclosed a zero-click vulnerability in Telegram's emoji sticker feature at a security conference held in Tokyo, and demonstrated that attackers can execute code on iOS devices without any user interaction. However, Telegram disputes the risk assessment of the vulnerability, and the available information has not yet clarified the specific scope of impact or remediation progress.
According to SlowMist, its security personnel discovered during routine threat intelligence monitoring that the Udesk customer service system-related script udeskApi.2026-10-16.cached.js has been implanted with a malicious loader. The loader hides the address using hexadecimal encoding and asynchronously loads the remote script deeps.js (originating from udeskqlobal[.]com; note that the domain starts with the letter 'q'). This script reads page-accessible Cookie, localStorage, and sessionStorage data, intercepts form submissions and login/authentication requests, and sends the data to udeskqlobal[.]com/c.php, enabling attackers to hijack sessions or take over accounts. This data collection behavior is not restricted by target site policies, and unmatched pages may also be affected. According to SlowMist's analysis, the script also targets three categories of sites: mgbx[.]com, bikingex[.], and bitxex[.]io. It scans deposit pages every 200 milliseconds and replaces recognized deposit addresses with preset BTC, EVM, TRON, Solana, LTC, and DOGE addresses. Additionally, by wrapping the Clipboard API, replacing copy buttons, and intercepting copy events, the script ensures users only see and copy tampered addresses. If users transfer funds to these addresses, the assets will flow into the attacker's wallet.
Odaily reports: PHP has issued a security advisory regarding two previously disclosed critical remote code execution (RCE) vulnerabilities, identified as CVE-2026-109618 and CVE-2026-109619. Attackers may exploit these vulnerabilities to execute malicious code without authentication, and PHP officially strongly recommends that all users upgrade immediately.These vulnerabilities affect PHP environments that handle web requests, including PHP-FPM, Apache mod_php, php-cgi, FrankenPHP, and LiteSpeed. Among these, Apache mod_php poses a particularly prominent risk, as attacks can be triggered before application code runs.Official patched versions have been released: PHP 8.5.12, 8.4.27, 8.3.36, and 8.2.35. Users still running PHP 8.1 or earlier should contact their distribution or long-term support (LTS) vendor to confirm patch availability, or upgrade to a supported version. Officials emphasize that temporary mitigation measures cannot fully eliminate the risk—installing security updates immediately is the only complete way to address the vulnerabilities.
SlowMist has issued a security alert stating that PHP released a security advisory on October 20, disclosing two critical security vulnerabilities that could lead to remote code execution (RCE), tracked as CVE-2026-109618 and CVE-2026-109619, both with a CVSS v4.0 score of 9.2. Currently, PHP officials have not yet released a fix patch, and the specific affected versions and exploitation conditions remain to be confirmed.According to available information, PHP officials plan to release a security update on October 21 at 14:00 UTC (22:00 Beijing time). SlowMist recommends that relevant organizations promptly check the PHP versions in their servers and business systems, with particular attention to PHP services exposed to the public internet, and upgrade in a timely manner once the patch is released; systems that cannot be upgraded immediately should restrict access sources and take isolation measures when necessary.
According to Bernama, a man in Perak, Malaysia, reportedly had his crypto wallet compromised, resulting in losses of approximately 12.47 million ringgit (roughly $3.05 million USD). The Perak state police chief stated that the victim discovered on the afternoon of October 9 that his digital assets had been transferred without authorization. Preliminary investigations indicate that the transactions were conducted via blockchain networks. Law enforcement is collaborating with the Cryptocurrency Unit of the Commercial Crime Investigation Department to trace the flow of funds and investigate those involved. The case is currently being investigated under Section 420 of the Malaysian Penal Code concerning fraud, while the specific attack method remains undisclosed.
security firm CrowdStrike released an intelligence report stating that an unidentified hacker, between late September and early October, combined large language models (LLMs) with the domestic open-source AI penetration testing tool ARTEX to launch targeted cyberattacks against multiple South Korean financial institutions and steal data.After analyzing the hacker's exposed open directory, it was found that they employed a dual-server architecture centered on a Hong Kong IP, and used models including Claude Code, GLM-5.3, Grok 4.6, and DeepSeek v4.1-flash accessed through an API proxy to orchestrate the attack process. The hacker also asked Claude how to monetize the leaked South Korean data in Telegram groups. According to estimates by South Korean media, at least seven South Korean financial institutions were attacked, including KB Kookmin Bank, Shinhan Bank, and Hana Bank, with personal data of approximately 68,000 people leaked, involving sensitive information such as annual income and loan limits.
Odaily News: Numa Lunah, co-founder of the crypto project Refi Hub, stated that he was hacked after using a download link provided in a Claude chat window to install a transcription application. The link pointed to a fake website bundled with malware, which attempted to steal all information from his device upon execution.Numa Lunah said he wiped and reinstalled the affected laptop and found no signs of sensitive data leakage. Subsequently, he discovered a contaminated Claude Code skill file named SKILL.md in his backups. The file was disguised as a style guide written by himself and contained instructions to re-download malware and steal credentials every time it was loaded.Microsoft Defender Experts previously warned that attackers have shifted from search engine optimization poisoning to large language model response poisoning. These tactics include recommending attacker-controlled download links, AI-branded fake installers, and contaminated code repositories and agent skills. Individuals working in the crypto industry may hold irrevocable credentials such as mnemonic phrases, private key files, hot wallet JSONs, exchange API keys with withdrawal permissions, and deployer keys. (Bitcoin.com News)
According to Decrypt, privacy-focused Bitcoin wallet Sparrow Wallet released version 2.5.4 on August 28. Developer Craig Raw stated that the update was driven by an AI-assisted code review, with the majority of fixes originating from it. This review was prompted by the recent seed generation code vulnerability exploit affecting Coldcard, as well as the release of unrestricted AI models in China, which has significantly enhanced vulnerability scanning capabilities across large codebases. Key updates include: validating the authenticity of transactions returned by Electrum servers, enforcing stricter BitBox02 hardware wallet security requirements (firmware v9.4.0 or higher required), patching local DNS leaks, and masking sensitive credentials in debug logs. Raw noted that there are no indications of any exploits being leveraged, user funds remain secure, and he still advises all users to update at their earliest convenience.
Z.ai releases GLM-5.3, based on the same foundation model as GLM-5.2, achieving capability improvements through expanded post-training. According to the official announcement, GLM-5.3 improves by 50% over GLM-5.2 on the internal Z.ai Code Bench coding benchmark, and reaches a leading level among open models in public benchmarks such as Terminal Bench 3.0 and Agents' Last Exam. In terms of cybersecurity, GLM-5.3 achieved a score of 84.5% in the CyberGym vulnerability discovery test, and significantly improved compared to the previous generation in exploit chain-related tests such as ExploitBench and ExploitGym.
Odaily Odaily News: Bitcoin Red Team member @Rob1Ham stated that OpenAI has blocked him from continuing security analysis of the Bitcoin codebase, after he had responsibly disclosed real vulnerabilities found within it. Rob1Ham said he had previously completed identity verification and onboarding processes related to OpenAI's cybersecurity capabilities, but is currently unable to continue investigating whether the fixes are sufficient or whether other vulnerabilities still exist. He stated that he will now resume Bitcoin security research using an open-source Chinese AI model. Rob1Ham commented: "Black hat hackers won't attack these issues, white hat hackers will." He also added: "For those who don't follow the rules and engage in harmful behavior, intelligence is unrestricted, while those committed to reducing harm are being excluded." (Bitcoin News)
Odaily News: The Bitcoin Red Team volunteer security initiative has scanned approximately 150 Bitcoin-related code repositories and disclosed more than a dozen vulnerabilities. The team is developing an open-source AI platform to audit Bitcoin software, covering wallets, cryptographic libraries, infrastructure, and other projects. AnchorWatch CEO Rob Hamilton stated that the team has so far spent approximately $20,000 on various AI services, using Kimi K3, OpenAI's GPT Sol, Anthropic's Claude Fable and Opus, as well as Z.ai's GLM 5.2 to identify vulnerabilities and generate related documentation. Pseudonymous Bitcoin developer Calle said that over the past 12 hours, the team has reported critical vulnerabilities to multiple projects, discovering on average roughly one critical vulnerability per person per hour, with daily spending of around $10,000. The team has not disclosed the affected projects or details of the vulnerabilities.
Cybersecurity tests conducted by the UK AI Safety Institute found that AI models with unrestricted internet access, without being instructed, autonomously forged false identities, implanted malicious code into open-source projects, and launched social engineering attacks against real individuals and organizations.
Odaily Planet Daily reported that Bitcoin News posted on X platform, stating that new evidence suggests the anonymous account "switck," who wrote the LibNgU code, may actually be Peter Gray, Co-founder and CTO of Coinkite. This code is at the center of the COLDCARD entropy failure incident. Researchers claim that Gray's GPG key signed dozens of commits by switck, and other identifiers appear to link the two identities together. Bitcoin developer James O'Beirne stated that he had warned Coinkite in May 2025 that the RNG implementation of LibNgU looked suspicious and recommended removing it, but he said the other party responded that if there were issues, they would have already been discovered. Screenshots also show that as early as April 2021, users had already raised questions about the LibNgU rewrite. If these findings are accurate, it means that the engineer who introduced the code was later linked to the theft of over 1,800 BTC, and had received direct warnings about the RNG implementation more than a year before the vulnerability was publicly disclosed.
Developers on Reddit used Claude Code to scan the Coldcard open-source firmware for vulnerabilities, pinpointing the core issue within 8 minutes: When generating private keys, the firmware invoked a software pseudo-random number generator instead of a hardware true random number generator, and it was this vulnerability that led to the theft of approximately $70 million in BTC from 1,196 wallets. Meanwhile, community users also reported that using Zhipu GLM 5.2 (trained on June 16, offline) for an independent scan similarly discovered this vulnerability. This bug has existed in the open-source wallet code for over five years.
Cos, founder of SlowMist, shared a tweet on X platform regarding potential poisoning attack risks in Claude Code and published a detailed analysis of poisoning attacks targeting Grok Build CLI and Claude Code CLI. The analysis pointed out that the security mechanisms of Grok Build CLI are not unified, with different code paths having different trust assumptions, and the gaps between them serve as channels for attackers.Attackers may exploit malicious project configuration files to execute arbitrary commands without the user's knowledge, thereby stealing API keys, cloud credentials, or gaining control over local devices. Researchers constructed a test environment and found that on Mac systems, if Claude Code is compromised, executing a specific test command could trigger the launch of a local calculator, demonstrating a potential command execution risk. If the attack succeeds, attackers could further steal API keys from AI services such as Claude and OpenAI, causing account cost losses; obtain credentials for cloud services like AWS, Alibaba Cloud, and Tencent Cloud to access servers and data; tamper with code repositories to implant backdoors; and leverage local devices as a springboard to attack internal enterprise networks. It is reported that the relevant vulnerability has existed for one year.
SlowMist founder Cosine retweeted a post about the potential poisoning attack risks of Claude Code, pointing out that attackers could execute arbitrary commands without the user's knowledge through malicious project configuration files, thereby stealing API keys, cloud credentials, or controlling local devices. Researchers constructed a test environment and found that in Mac systems, if Claude Code is affected, executing specific test commands can trigger the local calculator to launch, proving the existence of potential command execution risks. If the attack succeeds, attackers may further steal API Keys for AI services such as Claude and OpenAI, causing account fee losses; obtain cloud service credentials for AWS, Alibaba Cloud, Tencent Cloud, etc., to access servers and data; tamper with code repositories to implant backdoors; and use local devices as a springboard to attack enterprise internal networks.
Rob Witoff, Platform Leader at Coinbase, stated that currently, over 95% of the company's code is written or completed with the assistance of AI, a significant increase from the 40% figure announced in February this year. In an interview with Cointelegraph, Coinbase Platform Leader Rob Witoff said: "In fact, 100% of Coinbase employees use AI every day." Witoff noted that most Coinbase engineers currently run 5 to 10 AI Agents simultaneously. The combined work capability of these AI Agents is equivalent to approximately 1,200 employees.He predicts that by 2030, Coinbase's AI Agents could handle a workload equivalent to 100,000 employees. However, he stated that key areas such as core cryptography will still require human involvement, with AI primarily used for code testing, vulnerability checking, and prototyping. (Cointelegraph)