Suspected Chinese actor roots PAN-OS firewalls via chained PHP2Shell
The chain rooted 25 servers in four countries and reached a US government Active Directory domain, Unit 42 says.
86
- Mentions · 7d7
- Mentions · 30d15
- All mentions27
- First seenSep 1, 2026
- Last mentioned13h ago
Timeline
27 articles · newest firstAttackers chain PHP2Shell with Fragnesia kernel bug for root
Dream Security says the chain rooted Palo Alto firewalls and reached Active Directory on 25 servers, three tied to the US government.
84
Reports say PHP2Shell bugs were exploited, including on firewalls
GreyNoise reports at least six successful hits; Dream claims 24 firewalls and two US-linked servers were compromised.
84
Suspected Chinese attackers rooted 25 servers with custom PHP2Shell chain
The chain hit PAN-OS 12.1 and 12.2 from October 23; Dream says remote access likely remains.
84
Chained PHP2Shell exploit roots 25 firewalls, including US government systems
Dream says a custom PHP2Shell build plus the Fragnesia kernel bug yielded unauthenticated root and Active Directory access.
78
Attackers chain PHP2Shell and kernel exploits for root on 25 servers
Dream says custom exploits for CVE-2026-109618, CVE-2026-109619 and CVE-2026-46300 reached government-affiliated firewalls and Active Directory.
76
PHP2Shell campaign roots 25 servers, including U.S. government-linked hosts
Dream says a custom exploit plus a Linux kernel bug rooted firewalls in the U.S., Japan, Taiwan, and South Korea.
86
Criminals used Claude, Grok and Kimi against Latin American finance targets
Straiker says more than 400 organizations were targeted, not that each one was breached.
74
CyberCartel used Claude and Grok against 400-plus financial firms
Straiker reviewed the group’s CARLIY0S system and a snapshot of 4,493 records, without confirming every target was breached.
76
Chinese AI models parrot state doctrine or refuse to answer on sensitive topics
Aleph Alpha found Chinese models often repeat state doctrine or refuse questions on Tiananmen, Taiwan, and Xinjiang.
54
OpenAI Blocks 15,000 Requests Trying to Extract Protected Model Reasoning
OpenAI blocked over 15,000 accounts trying to extract protected model reasoning, linking much of the activity to Moonshot AI.
72
OpenAI Blocks 15,000-User Campaign Trying to Extract Hidden AI Reasoning
OpenAI stopped a 15,000-user campaign, partly tied to Moonshot AI, extracting hidden reasoning.
74
Academia is for Ambition — Alex Zhang, MIT
MIT's Alex Zhang discusses Recursive Language Models, GPU kernels, and multi-agent harnesses.
44
OpenAI says it stopped a campaign to steal its models' reasoning, but the trick still worked on Azure
OpenAI blocked a large campaign to steal model reasoning, but researchers say Azure still leaked it.
78
OpenAI Disrupts Reasoning Extraction Campaign Linked to Moonshot AI Associates
OpenAI disrupted a scaled reasoning-extraction campaign it links to people associated with China's Moonshot AI.
76
OpenAI reveals ‘novel’ encryption bypass used in distillation attack
OpenAI disrupted a Moonshot AI-linked distillation campaign that tricked its models into decrypting protected reasoning data using up to 15,000 accounts.
68
Irony alert: OpenAI whines that Chinese model stole its special IP that it stole from everybody else
OpenAI says it disrupted a July distillation campaign tied to Moonshot AI that copied protected model reasoning.
67
Disrupting a coordinated model-distillation campaign
OpenAI disrupted a July campaign extracting protected model reasoning, linking a core cluster to Moonshot AI.
78
LeapQuant: Efficient Linear Attention with Accurate Recurrent State Quantization
LeapQuant nearly matches FP32 quality with training-free 8-bit recurrent-state quantization.
44
Social Engineering in the Age of Synthetic Media
Recorded Future details how threat actors abuse jailbroken and legitimate AI tools to scale phishing, deepfake impersonation, and phishing-as-a-service operations.
55
Allspark: Weak to Strong Transfer via Alternating Chain of Thought
Allspark transfers weak-model reasoning to stronger students via alternating chains of thought, including Kimi and Nemotron.
55
Autonomous AI Agents Hack Retailers for $25 and Steal 600,000 Credit Cards
Gambit Security details an unsupervised AI-agent campaign using Strix, Cairn and Hermes that breached 27+ retailers and stole 600,000+ credit cards for about $25 per target.
78
OpenRouter's staggering token chart is the AI bubble debate in a single image
OpenRouter weekly token consumption surged 25,000% since January 2025, but reasoning-model 'thinking' tokens inflate the metric beyond real adoption.
32
Moonshot serves Claude instead of Kimi and collects exchanges for model training
Moonshot AI reportedly served users Anthropic's Claude instead of its own Kimi model and collected the exchanges for model training.
35
Anthropic details distillation campaigns from Alibaba, Moonshot AI, and DeepSeek
Anthropic reports nearly 200 million Claude exchanges tied to distillation campaigns by Alibaba, Moonshot AI, and DeepSeek.
78
Chinese AI firms are siphoning capabilities from American models, CISA warns
CISA, NSA and FBI warn Chinese AI firms including DeepSeek and Moonshot AI extracted billions of tokens from US frontier models via distillation campaigns.
70
Bilevel Coordinated Reflection: A Game-Theoretic Approach to Multi-Agent LLM Systems
Paper models multi-agent LLM orchestration as a bilevel game, proving transcript-only gating limits and introducing grounded-memory SRMA.
32
Related CVEs
by mentions- CVE-2026-109619100%9.2Use-after-free RCE in PHP via stale request globals
- CVE-2026-109618100%9.2OOM longjmp memory flaw enabling RCE in PHP
- CVE-2026-4630016%7.8Shared-frag marker loss in Linux kernel skb coalescing (LPE)
Appears with
- Moonshot AIvendor · 10
- Moonshot AIorganization · 10
- Anthropicvendor · 8
- DeepSeekai model · 8
- OpenAIorganization · 8
- OpenAIvendor · 7
- Dreamorganization · 6
- Frontier Model Forumorganization · 6
- Anthropicorganization · 6
- Claudeai model · 6
- CISAorganization · 4
- Linux kernelproduct · 4
- Qwenai model · 4
- PHPproduct · 4
- DeepSeekorganization · 4
- Kimiproduct · 4
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.