跳到主要内容

Security Daily Brief

20 events32 source observations

Lead stories

CVE-2026-82568 The Give - Peer-to-Peer plugin for WordPress is vulnerable to time-based SQL Injection via the 'sort' parameter in all versions up to, and including, 2.9.3 due to in

CVE-2026-82568 The Give - Peer-to-Peer plugin for WordPress is vulnerable to time-based SQL Injection via the 'sort' parameter in all versions up to, and including, 2.9.3 due to in

Read the analysis

The day in context

Today's security landscape is dominated by a wave of critical vulnerabilities affecting software ranging from WordPress plugins to industrial control systems. The Give-Peer-to-Peer plugin for WordPress faces two severe flaws: CVE-2026-82568 enables time-based SQL injection via the 'sort' parameter in versions up to 2.9.3, while CVE-2026-82663 allows authorization bypass in versions up to 2.9.5. These issues highlight a pattern of authentication and data integrity failures in widely used web components. Simultaneously, industrial and embedded systems are under scrutiny; a vulnerability in D-Link DSL-6740C routers and a flaw in Rexroth IndraWorks allow attackers to execute commands, underscoring risks in legacy hardware and automation software. Furthermore, Prime CMS and Cyrus SASL contain authentication bypass vulnerabilities, with the latter affecting SRP server mechanisms in versions through 2.1.28.

Beyond software flaws, active exploitation and new malware campaigns are reshaping the threat environment. Rapid7 detected exploitation of CVE-2026-90436 against Atlassian Confluence Data Center on November 13, 2026, indicating real-world impact. The Retadup USB worm has resurfaced seven years after its takedown to spread CoffeeLoader malware, demonstrating the persistence of known threats. In the AI-driven security space, Amazon uncovered an AI agent campaign targeting LiteLLM and Dify for API keys, while ZDNET reports suggest AI tools are now enabling mediocre hackers to find zero-day flaws for a low cost. This convergence of AI in both offense and defense creates a complex operating picture where traditional detection methods face new challenges.

The threat landscape also extends to data breaches and specialized malware. TSMC confirmed a data breach involving employee personal data linked to a cloud provider, raising concerns about supply chain security. Criminal marketplaces are now selling infostealer logs with AI-written profiles of victims, adding a layer of sophistication to data black markets. Additionally, ArcticRAT Android malware utilizes Grok AI to analyze stolen banking texts, showing how AI is integrated into financial theft tools. These events collectively illustrate a shift towards automated, AI-enhanced attacks targeting identity, infrastructure, and financial data across diverse sectors.

Lead stories

CVE-2026-82568 The Give - Peer-to-Peer plugin for WordPress is vulnerable to time-based SQL Injection via the 'sort' parameter in all versions up to, and including, 2.9.3 due to in

Published:

The Give Peer-to-Peer plugin for WordPress has been identified with a critical time-based SQL injection vulnerability affecting all versions up to and including 2.9.3. This flaw exists specifically within the sort parameter handling mechanism, where insufficient escaping of user-supplied input allows attackers to manipulate database queries. The vulnerability enables remote code execution by forcing the database to wait for a specific duration before responding, effectively draining server resources and exposing the underlying data structure to unauthorized manipulation. This issue impacts a wide range of e-commerce sites and donation platforms that rely on the Peer-to-Peer functionality for their core operations. Given the widespread adoption of WordPress in the security sector, this vulnerability poses a significant risk to the integrity of financial transactions and customer data across numerous organizations. Immediate patching is required to mitigate the potential for exploitation by malicious actors seeking to compromise these systems.

AI finds zero-day flaws for just $4 now - even mediocre hackers can cash in - ZDNET

Published:

Recent analysis indicates that artificial intelligence is increasingly capable of identifying previously unknown software vulnerabilities, with some models successfully locating private zero-day flaws in major systems like Google's V8 engine. This capability has lowered the financial barrier to entry for exploitation, allowing even less skilled actors to acquire access to critical vulnerabilities for as little as four dollars. The trend suggests a significant shift in the threat landscape where automated tools are democratizing the ability to discover and monetize security gaps that were previously reserved for elite researchers or state-sponsored groups. Consequently, defenders face a heightened challenge in keeping pace with these rapid discoveries, as the volume of potential threats may outstrip traditional detection and patching efforts. The emergence of such low-cost access points implies that organizations must reconsider their security posture, potentially prioritizing proactive measures over reactive defense strategies to mitigate the rising tide of automated exploitation.

Criminal marketplace sells infostealer logs with AI-written profiles of their owners - Help Net Security

Published:

A criminal marketplace recently advertised on cybercrime forums for selling data harvested from computers infected with infostealer malware. This illicit offering includes not only the raw logs of the infection but also AI-generated profiles specifically crafted to describe the apparent owners of the compromised systems. Hudson Rock discovered this specific listing, highlighting a new dimension to the data black market where stolen credentials are paired with synthetic biographical details. The immediate context involves the active circulation of high-value intelligence on dark web platforms where threat actors monetize their operations. This development matters to a professional reader because it demonstrates the evolving sophistication of criminal toolkits, moving beyond simple credential dumps to include automated, personalized dossiers that could facilitate more targeted social engineering or identity theft campaigns against victims of the infostealer malware.

Complete briefing

Vulnerabilities & patches

7

CVE-2026-82663 The Give - Peer-to-Peer plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 2.9.5.

Published:

CVE-2026-82663 The Give - Peer-to-Peer plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 2.9.5. This is due to the plugin not properly… cve.org/CVERecord?id=C…

nltk: 3 security advisories

Published:

GitHub published 3 security advisories affecting nltk. Reported issues include: NLTK dot2img runs Graphviz 'dot' without the trusted-exec chokepoint, enabling local code execution via a writable PATH element; NLTK: Symlink escape in chat80.sql_query bypasses the pathsec data-root sandbox to read out-of-root SQLite databases; NLTK: Path Traversal in save_maxent_params / load_maxent_params Allow Arbitrary File READ/WRITE.

CVE-2026-129444 A vulnerability was identified in D-Link DSL-6740C 20131126.

Published:

CVE-2026-129444 A vulnerability was identified in D-Link DSL-6740C 20131126. This affects an unknown function of the file /etc/templates/ftpd/ftpd_conf.php of the component FTP Dae… cve.org/CVERecord?id=C…

Rapid7 Detected Exploitation of CVE-2026-90436 against Atlassian Confluence Data Center, N-able N-central also Exploited

Published:

On November 13, 2026, Rapid7 MDR detected exploitation of CVE-2026-90436 against Atlassian Confluence Data Center in a customer environment. Separately, on November 13, 2026, N-able reported exploitation in the wild against N-central for a vulnerability that strongly correlates to CVE-2026-90436.

A vulnerability has been identified in Rexroth IndraWorks. This flaw allows an attacker to...

Published:

A vulnerability has been identified in Rexroth IndraWorks. This flaw allows an attacker to execute arbitrary code on the user’s system by parsing a manipulated file, because the file parser fails to properly sanitize the path name(s) or references embedded in the manipulated file.

CVE-2026-129941 Prime CMS (@primecms/core) through 0.4.0-beta.0 contains an authentication bypass vulnerability that allows unauthenticated attackers to register accounts via the c

Published:

CVE-2026-129941 Prime CMS (@primecms/core) through 0.4.0-beta.0 contains an authentication bypass vulnerability that allows unauthenticated attackers to register accounts via the c

CVE-2026-129483 Cyrus SASL through 2.1.28 contains an authentication bypass vulnerability in the SRP server mechanism because srp_server_mech_step2 does not reject client public va

Published:

CVE-2026-129483 Cyrus SASL through 2.1.28 contains an authentication bypass vulnerability in the SRP server mechanism because srp_server_mech_step2 does not reject client public va

Exploitation & attacks

3

TSMC employee data breach linked to cloud provider

Published:

Taiwan Semiconductor Manufacturing Company (TSMC) confirmed that personal employee data has been leaked online after a threat actor put 6.9GB of information allegedly stolen from the company up for auction.

Amazon Uncovers AI Agent Campaign Targeting LiteLLM, Dify, and Other Apps for API Keys

Published:

Attackers' AI agents have been hunting for weak spots in internet-exposed AI apps such as LiteLLM, One API, and Dify to steal the AI provider keys stored inside, according to the agents' own records.

Malware

3

Research & tools

2

Identity & phishing

2