Cybersecurity
Every day we read hundreds of stories. Only the most important ones are here — short, in our own words, with a link to the original source.
- Cybersecurity
JetBrains responds to eMudhra CA security incident
JetBrains is addressing an industry-wide security incident at certificate authority eMudhra that affects certificate trust in several of its products. The company advises customers to keep their products and systems updated while fixes are released.
Full story at blog.jetbrains.com ↗ - Artificial intelligence · Cybersecurity
OpenAI monitors AI models for misalignment
OpenAI uses layered monitoring to detect frontier model misalignment and unauthorized agent behavior. The system aims to let responders intervene before incidents escalate.
Full story at openai.com ↗ - Cybersecurity · Artificial intelligence
Voice prompts can bypass AI safety checks
Check Point researchers found that varying speed, pitch and background noise got a voice AI model to give harmful answers in about 70 percent of attempts. Text-based transcript checks often fail to catch such audio manipulation.
Full story at blog.checkpoint.com ↗ - Cybersecurity · Artificial intelligence · New AI models · Open source
Open model GLM 5.3 builds Firefox exploits on its own
Epoch AI says the widely available GLM 5.3 model turned public security fixes into working exploits before updates reached Firefox users. It autonomously compromised six Firefox releases for about $1,000 in API fees each, without triggering safeguards from any provider tested.
Full story at The Decoder ↗ - Cybersecurity
Hackers drain billions in crypto from Binance
Binance suspended withdrawals after hackers reportedly stole $6.3 billion in digital assets. TRM Labs has high confidence North Korea's elite hacking unit was behind it, and Binance says it froze about $3 billion.
Full story at Futurism ↗ - Cybersecurity · Programming & dev tools · Artificial intelligence
Tampered Claude Code add-on found in Cursor marketplace
A tampered Claude Code extension was live on the Cursor marketplace for about 11 hours, affecting up to 5,000 users. It tried to steal data and set up remote access; Cursor took the marketplace offline and restored it after confirming no further risk.
Full story at The Decoder ↗ - Cybersecurity · Programming & dev tools · Open source
Node.js patches serious Brotli decoder flaw
On October 21, 2026, the Node.js project released updates fixing CVE-2026-111497, an out-of-bounds write in the Brotli decoder. Any application decompressing attacker-controlled data with a vulnerable Brotli library could be at risk; Google scored it 9.2 under CVSS v4.0.
Full story at sentinelone.com ↗ - Cybersecurity · Smartphones & apps
National CERT warns of spyware delivered through ads
National CERT says a phone can be infected with spyware just by loading an online ad. No click is needed, and attackers misuse advertising networks in a technique called "AdFection".
Full story at ProPakistani ↗ - Cybersecurity · Networking
Cloudflare tracks abuse of eMudhra's weak keys
On October 19, 2026 eight extremely weak eMudhra keys became known, one of them simply the number 4. Cloudflare now shares evidence of exploitation and progress on patching.
Full story at blog.cloudflare.com ↗ - Cybersecurity · Smartphones & apps · Operating systems
A text message can brick phones, and Google has a fix
The so-called SMS of Death can brick a phone. Google has already issued a fix, according to the report.
Full story at PhoneArena ↗
An AI writes the short texts from the source's stories.
Free by e-mail
Choose your topics. You only get the most important news on them.