Topic
Everything filed under Security, newest first.
RSS · JSON · All topics
A rebuild could clear 28% of our client's container CVEs. The rest needed evidence.
A red container-vulnerability test in the middle of a SOC 2 observation window. A rebuild could clear 28% of the findings. The rest took scanners that never agreed, VEX claims customers act on, risk acceptances with an owner and a date, and a reachability test that proved my own argument wrong.
14 min · 3,246 words
Pulling the right strings for RCE on a Puppet
Neodyme describes two critical CVSS 9+ vulnerabilities found in Puppet ahead of a red team engagement: a remote code execution on the Puppet server in default configuration (CVE-2026-115731) and a host impersonation flaw in a common non-default setup (CVE-2026-115732), plus disclosure details.
10 min · 2,239 words
So my ISP was hijacking my DNS
Karthikey Hegde found his ISP tampering with plaintext DNS queries and returning bogus IP addresses, then explains common misconceptions about DoH, DoT and DNSSEC and describes the AdGuard plus dnscrypt-proxy setup he now runs in his homelab.
4 min · 902 words
Yannick Dixken details the vulnerabilities, including remote code execution, he found in Prusa's Buddy 3D-printer firmware, his AI-assisted method, the coordinated disclosure with Prusa (fixed above firmware 7.0.0), and why such research got cheap.
20 min · 4,639 words
PHP2Shell and AI Put Half the Internet on Edge
IEEE Spectrum's Dina Genkina reports on the PHP2Shell vulnerability and argues AI is forcing the industry to pay down decades of technical debt in legacy software that powers much of the web.
5 min · 1,083 words
An angry post arguing that Ubuntu failed its promise of timely patches: more than two weeks after Debian patched the critical remote Linux TCP bug CVE-2026-98432, every supported Ubuntu release was still listed as vulnerable, so the author took his Ubuntu servers offline and is moving off the distro.
1 min · 301 words
Daniil Gentili announces Psalm 7, the PHP static analysis tool's biggest release: security taint analysis on by default with 10x+ speedups and deeper coverage (NoSQL, LLM prompt and URL injection), a capability-based purity model, graph-based dead code detection, template defaults, PHP 8.5/8.6 support and daily releases.
11 min · 2,447 words
Progress on provable inference for large language models
Anthropic's Frontier Red Team and collaborators from EPFL, MIT and elsewhere describe progress on cryptographic proofs that an LLM's output token stream really came from a specified model, including a prototype system, its performance costs, and open problems for verifiable inference.
21 min · 4,805 words
Jailbreaking my car's navigation system just so I can change the time
Rasmus Moorats reverse engineers the navigation head unit of his 2014 Renault Latitude, which oddly needs a map SD card just to change the clock, and ends up patching the timezone offset in a custom binary file to set UTC+2.
11 min · 2,633 words
The Rise of LLMJacking: How Threat Actors Monetize Compromised Cloud and Model Resources
Google Threat Intelligence Group and Mandiant describe 2026 incidents where attackers hijacked enterprise LLM access across cloud providers, running up fraudulent charges above $100,000 in a day and up to $1 million cumulatively, plus how the access is resold, and detection, hunting and mitigation guidance with IOCs.
13 min · 3,080 words
so i guess everyone uses llms to think now
Security researcher eva explains why she never uses AI for the research on her blog: tinkering is the fun, she can't take credit for an LLM's work, and she argues AI firms use hyped vulnerability disclosures, PoC videos and hash drops (php2shell, 'tailworm') as marketing.
8 min · 1,760 words
Security Disclosure: Sunday November 1st
Trigger.dev CTO Eric Allam discloses a security incident in which an attacker gained unauthorized access to environment variables for a subset of customer organizations; the exploit has been blocked and impacted account owners were emailed mitigation steps.
1 min · 67 words
It wasn't AI that flipped Proton's charge
Longtime Proton customer Raed Chammam argues the backlash over Proton's AI push is really about leadership: a CEO calling an engineer's request for pentesting time "loser thinking" undermines the careful engineering a privacy company sells, regardless of whether code is AI-written.
2 min · 456 words
Colin Percival recounts finding in 2004 that Intel Hyper-Threading leaked information between threads via cache state, spending six months on coordinated disclosure before publishing in 2005, and why he sees a scholar's duty as sharing knowledge with the people who can act on it.
7 min · 1,717 words
George Hotz argues AI can play out as either order or chaos, that order (models patching every exploit and locking in the status quo) is the dangerous outcome, and that chaos, with botted media, cyberattacks and forced decentralisation, is the better one.
2 min · 530 words
Appraisal by AI: Model-driven victim prioritization in the VOLTA infostealer
Cisco Talos analyses VOLTA, a self-deleting infostealer that, after stealing credentials, sessions and crypto wallets, sends a structured evidence bundle to an LLM to rank the victim's assets by monetization value, failing over across a three-model chain when a model refuses.
7 min · 1,556 words
Inside a malicious Claude Code-branded extension and its persistent SSH backdoor
Microsoft Security Research analyzes the tampered Claude Code extension briefly served on the Cursor extension marketplace in October 2026, detailing its host reconnaissance, reverse SSH tunneling backdoor and startup persistence, plus detection, remediation guidance and IOCs.
12 min · 2,698 words
Introducing sanstun: A sans-IO, parallel-friendly, userspace WireGuard implementation in Rust
Firezone's Thomas Eizinger announces sanstun, an open-source Rust implementation of the WireGuard state machine that separates IO from protocol logic, and explains its sans-IO design, how its API lets encryption and decryption run in parallel while the interface stays single-threaded, and why it replaces Firezone's boringtun fork.
8 min · 1,848 words
Getting Google Cast on my Soundbar
Matthew Garrett pulls the firmware off a Samsung Q800A soundbar, gets a root shell via Samsung's sdb debug bridge and the WAM protocol, and discovers that Google Cast support is gated by country code, then flips it on by making the device think it is Australian.
6 min · 1,312 words
4 real? Tracking exploitation and patching of eMudhra’s weak keys
Cloudflare researchers follow up on certificate authority eMudhra's eight trivially weak intermediate keys, which anyone could recover from the public certificates, covering how Cloudflare patched, whether exploitation was seen across its network, how many visitors remain vulnerable, what users should do, and why trust stores need fast revocation.
7 min · 1,580 words