The CTO of RansomwareHelp, a site that claimed its cryptographic tools could fully restore ransomware-encrypted systems and urged victims not to pay, has been indicted for wire fraud. Reporting by 404 Media and Brian Krebs examined the indictment and the site's history.
The CTO and co-founder of ransomwarehelp.com, a Florida firm claiming to restore ransomware-locked data with advanced cryptographic tools, was indicted for wire fraud. An FBI sting found the company paid attackers for decryption keys and charged victims a markup for work never performed.
An article reports that cybersecurity leaders are treating clear, disciplined communication, from explaining cloud security to guiding AI use, as a front-line defense capability across the enterprise. It argues that communication now matters as much as, or more than, tooling.
Researchers at the University of Waterloo, Zhejiang University and KTH Royal Institute of Technology found AI agents retrained on larger prompt-injection datasets were easier to hijack in three of four tests. The automated red-teaming tool that found more working attacks produced a weaker defense.
ToolJet, an open-source low-code internal tool builder, released version 4.0 with a rebuilt app builder. The founder said customers wanted AI-assisted "vibe coding" flexibility while retaining the governance controls ToolJet provides, and that the product was rebuilt rather than repositioned.
BackBox Linux released version 10, codenamed Resolute Raccoon, on November 19. It is built for Ubuntu 26.04 LTS compatibility, ships the Linux 7.0 kernel and Xfce 4.20, and now routes DNS through Tor in anonymous mode while blocking IPv6.
Dashlane research found that about one in five UK employees share passwords with AI tools and grant those tools access to sensitive data. The report frames this as an enterprise security risk as generative AI use spreads through workplaces.
The US Homeland Security Task Force launched a Threat Disruption Center within its National Coordination Center in Fairfax, Virginia. The unit will give investigators tools to triage victim complaints and map foreign scam, cybercrime and sextortion networks, and will work with private firms to intercept victims' money before it reaches them.
Apache OpenNLP's built-in regex name finder for latitude and longitude coordinates uses a pattern vulnerable to excessive backtracking. Text supplied to that finder can trigger a denial of service; opennlp-tools 2.0.0 before 2.5.13 and opennlp-runtime 3.0.0-M1 before 3.0.0 are affected.
Nudge Security's AI Adoption in Practice 2026 report, based on anonymized telemetry from customer environments, says AI agents built inside project trackers, CRMs and automation tools typically run on their creator's credentials. Those agents therefore inherit the employee's data access, and organizations rarely monitor what they do with it.
A suspected Chinese APT targeting Philippine government organizations has rebuilt its EggStreme malware toolkit, with the new variant going dormant to evade security tools. Bitdefender said the attackers rotated infrastructure and files after exposure, changing file hashes on each recompile.
A Lobste.rs discussion cites reports that Windows 11 gathers telemetry about which coding agents and developer tools users run. The claim is based on a single source and no technical detail or vendor confirmation is included.
Detectify released Agentic Tooling, which lets customers' own AI agents run in their environments to investigate suspected vulnerabilities and test exploitability in live applications. A companion Security Ledger lets security teams check agents' findings against objective evidence rather than probabilistic conclusions.
Google's Threat Intelligence Group reported that Russian-speaking hackers conducted what it assesses as the first cyberattack run end-to-end by artificial intelligence without human involvement at any stage. The tooling trades stealth for speed, leaving noisy forensic traces even when intrusions succeed.
Coral, an open-source self-hosted gateway, was launched on Hacker News by co-founder James. It sits between AI agents and target systems, letting agents author and save short TypeScript programs as reusable tools that can be published to groups of agents for permission control.
A Reddit security post titled "Top Command and Control Tools of 2026" was submitted by a single user and aggregated by one source. It lists C2 tools but the snippet contains no details or vendor information.
Security Affairs published a newsletter roundup of AI cybersecurity articles and research covering the international landscape. It says AI agents can automate tasks, analyze large data sets, discover vulnerabilities, and accelerate offensive operations.
A Reddit security post describes a campaign using fake Google Meet update lures alongside abused RMM tools and an ecosystem of RAT payloads referred to as AgtaBackup. The post aggregates the activity from user submissions and comments.
Cisco Talos reports growing in-the-wild use of AI-enabled penetration testing frameworks, including ARTEX, which automate and chain reconnaissance, vulnerability testing, and follow-on actions. Talos says the tools lower the skill threshold for attackers and accelerate intrusions when misused.