Weekly Threat Intelligence Summary

Top 10 General Cyber Threats

Generated 2026-10-12T05:00:06.392631+00:00

1. Chinese Government-linked Cyber Threat Actors Combine Automated and Hands-on Hacking Tools to Steal Sensitive Data (www.cisa.gov, 2026-10-06T16:22:53)
   Score: 17.879
   Advisory at a Glance Title Chinese Government-linked Cyber Threat Actors Combine Automated and Hands-on Hacking Tools to Steal Sensitive Data Original Publication October 8, 2026 Executive Summary Chinese government-linked cyber threat actors, enabled by the Integrity Technology Group, are combining automated scanning tools, large-scale botnets, and hands-on exploitation techniques to target and steal sensitive data from organizations worldwide, including US critical infrastructure sectors. Thes
2. Google pauses open source bug bounty program after rise in AI submissions (www.malwarebytes.com, 2026-10-05T15:00:56)
   Score: 7.103
   Google has frozen its OSS VRP product vulnerability submissions to stop the flood of AI generated vulnerability reports.
3. CrowdStrike and Anthropic Give Critical Infrastructure Defenders the AI Advantage (www.crowdstrike.com, 2026-10-08T05:00:00)
   Score: 7.033
4. Japan Adopts Proactive Cyber Defense Strategy (www.recordedfuture.com, 2026-10-08T00:00:00)
   Score: 6.999
   Learn about Japan's Active Cyber Defense (ACD) strategy, mandatory reporting rules, and key impacts on critical infrastructure.
5. Update Chrome and ChromeOS to fix critical security issues (www.malwarebytes.com, 2026-10-07T10:25:26)
   Score: 6.904
   Google has released a new update for the Chrome browser and ChromeOS. There are a lot of security fixes in them so don't delay.
6. Malwarebytes earns another Top Product award in independent testing (www.malwarebytes.com, 2026-10-01T10:51:52)
   Score: 6.407
   Three independent labs, three standout results: a perfect score, top certification, and every threat stopped before it ran.
7. Using Threat Intelligence to Stop Ransomware Attacks (www.recordedfuture.com, 2026-09-25T00:00:00)
   Score: 6.332
   Learn how ransomware threat intelligence empowers your team to actively follow adversary infrastructure, monitor dark web chatter and prevent attacks.
8. Tajin Group: Guarantee Marketplace Vendor Involved in Phishing and Chinese Money Laundering Group (www.recordedfuture.com, 2026-09-15T00:00:00)
   Score: 6.165
   Analyze Tajin Group's role in phishing and Chinese money laundering. Discover how this Telegram-based vendor exploits payment gateways and adapts its financial fraud operations.
9. ASOS breach update: Hackers stole customer details and shopping searches (www.malwarebytes.com, 2026-10-09T10:56:41)
   Score: 5.741
   Stolen ASOS data includes shopping searches as well as personal details, so customers should watch out for convincing phishing messages.
10. Recorded Future Introduces AI Infrastructure Indicator Lists, Strengthening AI Governance (www.recordedfuture.com, 2026-10-09T00:00:00)
    Score: 5.665
    Recorded Future's new AI Infrastructure Indicator Lists help security teams find shadow AI, enforce policy, and prevent data loss.

Top 10 AI / LLM-Related Threats

Generated 2026-10-12T06:00:20.244578+00:00

1. Cancerous Backdoor Attack: Gradual Activation via Invisible Tokens and A Stage Controller (arxiv.org, 2026-10-12T04:00:00)
   Score: 17.78
   arXiv:2610.12644v1 Announce Type: new
   Abstract: Backdoor Attacks on Large Language Models (LLMs) typically follow a trigger-response pattern, where satisfying a trigger condition activates a single terminal behavior. However, this single-point backdoor activation does not explore attacks in which malicious behavior develops gradually through multiple intermediate stages before reaching the terminal activation stage. We present Cancerous Backdoor Attack, that gradually corrupts an LLM's rea
2. EviRepair: Evidence-Driven Agents for Reliable Vulnerability Repair (arxiv.org, 2026-10-12T04:00:00)
   Score: 17.78
   arXiv:2610.13491v1 Announce Type: new
   Abstract: Recent advances in large language models (LLMs) have enabled agent-based approaches for repository-level automated vulnerability repair (AVR), showing promising repair capabilities. However, an important issue has received limited attention: how can we ensure that the evidence generated during the repair process is reliable enough to support accurate vulnerability understanding and patch generation? In this paper, we propose EviRepair, an evidence
3. Byzantine-Resilient Federated Large Language Model via Hypergraph Message Passing (arxiv.org, 2026-10-12T04:00:00)
   Score: 17.78
   arXiv:2610.12711v1 Announce Type: cross
   Abstract: Federated Large Language Models (FedLLMs) are promising to deliver privacy-preserving, collaborative intelligence. The decentralized training process is vulnerable to Byzantine attacks that corrupt semantic consistency and factual resilience through malicious model updates. This paper presents a new resilience strategy that prevents semantically corrupted knowledge from propagating during federated aggregation. A defense problem is formulated to
4. LLMail-Inject: A Dataset from a Realistic Adaptive Prompt Injection Challenge (arxiv.org, 2026-10-12T04:00:00)
   Score: 17.48
   arXiv:2506.09956v2 Announce Type: replace
   Abstract: Indirect prompt injection attacks exploit a fundamental weakness of LLMs: executing unauthorized tasks from untrusted sources. Despite the practical significance of this attack vector, existing benchmarks rely on small sets of template-based attacks (100-1,000 examples) that fail to capture the diversity and adaptiveness of real-world adversaries. Consequently, benchmarks are saturated and overfit to static attacks. We introduce LLMail-Inject,
5. Auditing Prompt Injection Defenses Before They Reach Users: What Strong Adversaries Reveal (arxiv.org, 2026-10-12T04:00:00)
   Score: 15.78
   arXiv:2505.15738v3 Announce Type: replace
   Abstract: Large language models are increasingly deployed in user-facing products that handle sensitive personal data while interacting with untrusted external content. This makes them a prime target for prompt injection attacks, in which an adversary injects malicious instructions in that content to hijack the model. Fine-tuning-based defenses have been proposed to mitigate these attacks by training models to follow benign user instructions and ignore
6. ToFu: Exploiting and Fuzzing the LLM-Orchestrator-Tool Trust Boundary (arxiv.org, 2026-10-12T04:00:00)
   Score: 14.78
   arXiv:2610.13651v1 Announce Type: new
   Abstract: Large language models (LLMs) are increasingly deployed as agents that interact with external software through tools, with an orchestration layer translating model outputs into structured tool invocations. This architecture creates an LLM-orchestrator-tool trust boundary where tool semantics may go unenforced and LLM-generated arguments may be incorrectly treated as trusted. We present ToFu, a tool-aware framework that automatically derives semanti
7. AutoControl Arena: Synthesizing Executable Test Environments for Frontier AI Risk Evaluation (arxiv.org, 2026-10-12T04:00:00)
   Score: 14.78
   arXiv:2603.07427v3 Announce Type: replace-cross
   Abstract: As Large Language Models (LLMs) evolve into autonomous agents, existing safety evaluations face a fundamental trade-off: manual benchmarks are costly, while LLM-based simulators are scalable but suffer from logic hallucination. We present AutoControl Arena, an automated framework for frontier AI risk evaluation built on the principle of logic-narrative decoupling. By grounding deterministic state in executable code while delegating gener
8. Pwn2Own Ireland 2026 – Day One Results (www.thezdi.com, 2026-10-06T09:28:31)
   Score: 13.106
   Welcome to Day One of Pwn2Own Ireland 2026! Today, 21 entries took the Pwn2Own stage to target multiple categories – including multiple Samsung attempts and our first Pixel entry. Follow the action live! We’ll be posting real-time updates and results throughout the competition on our blog and across social media. Stay up to date by following us on Twitter , Mastodon , LinkedIn , and Bluesky , and join the conversation using #Pwn2Own Ireland and #P2OIreland for continuous coverage. Here are the r
9. Pwn2Own Ireland 2026 – The Full Schedule (www.thezdi.com, 2026-10-05T18:11:29)
   Score: 12.954
   Fáilte! Pwn2Own Ireland 2026 has arrived in Cork, and the world's top security researchers are ready. We've got more phone entries than we've ever had, and more than 60 entries total. Earlier today, we held the random draw to determine attempt order. Below is the official schedule. All times are Irish Standard Time (IST) and may change as the competition progresses. You can find our live leaderboard here . This will be updated throughout the day. Jump to: — Day One — Day Two —
10. Understanding Verification Vulnerabilities in Cross-chain Bridges: An Empirical Study (arxiv.org, 2026-10-12T04:00:00)
    Score: 12.48
    arXiv:2610.13170v1 Announce Type: new
    Abstract: With the rapid development of blockchain technology, numerous heterogeneous blockchains have emerged, each featuring distinct architectures, consensus mechanisms, and application ecosystems. Cross-chain bridges have consequently become a critical infrastructure component, typically relying on smart contracts to facilitate message passing and asset transfers across heterogeneous blockchains. Unlike single-chain contracts, bridge contracts introduce
11. Securing Agent-to-Agent Communication: The Next Identity Frontier (www.rapid7.com, 2026-10-06T14:11:03)
    Score: 12.153
    As organizations deploy autonomous AI agents, security teams face a significant shift as non-human non-human entities making decisions, invoking tools, and delegating tasks to other agents without human intervention. Security architectures built around human users, static APIs, and distinct endpoints break down when AI agents dynamically collaborate across an environment. As these interactions become more common, securing agent-to-agent communication without blocking adoption will require securi
12. Evolution of Web3 in Cloud Supply Chain Attacks (unit42.paloaltonetworks.com, 2026-10-07T22:00:16)
    Score: 11.768
    Unit 42 details how threat actors leverage Web3 infrastructure and open-source supply chain attacks to breach enterprise cloud environments The post Evolution of Web3 in Cloud Supply Chain Attacks appeared first on Unit 42 .
13. Rethinking access control for RAG with Amazon Quick and Amazon Bedrock (aws.amazon.com, 2026-10-07T18:34:44)
    Score: 11.334
    Enterprise RAG unlocks insights from knowledge sources like SharePoint, Google Drive, and Confluence, but those sources carry complex permissions. Learn how Amazon Quick and Amazon Bedrock Knowledge Bases enforce document-level access controls in real time, verifying permissions directly with authoritative sources at query time.
14. Chinese Government-linked Cyber Threat Actors Combine Automated and Hands-on Hacking Tools to Steal Sensitive Data (www.cisa.gov, 2026-10-06T16:22:53)
    Score: 11.174
    Advisory at a Glance Title Chinese Government-linked Cyber Threat Actors Combine Automated and Hands-on Hacking Tools to Steal Sensitive Data Original Publication October 8, 2026 Executive Summary Chinese government-linked cyber threat actors, enabled by the Integrity Technology Group, are combining automated scanning tools, large-scale botnets, and hands-on exploitation techniques to target and steal sensitive data from organizations worldwide, including US critical infrastructure sectors. Thes
15. Vulnerable Models in the AI Supply Chain: An Empirical Study of Adoption, Propagation, and Usage Patterns (arxiv.org, 2026-10-12T04:00:00)
    Score: 10.98
    arXiv:2610.12893v1 Announce Type: new
    Abstract: The rapid integration of Artificial Intelligence (AI) models has established a complex supply chain, but many of these shared artifacts harbor severe vulnerabilities such as insecure serialization and architectural backdoors. While prior efforts have extensively scanned model registries to flag malicious artifacts, a critical gap remains in understanding how developers actually adopt and integrate these unsafe models into downstream software syste
16. OpenAI Fires 3 Safety Researchers in Dispute Over AI Risks (www.securityweek.com, 2026-10-09T21:07:46)
    Score: 10.936
    The ChatGPT maker said the researchers "violated clear policies on handling sensitive information.” The post OpenAI Fires 3 Safety Researchers in Dispute Over AI Risks appeared first on SecurityWeek .
17. An Exploration of Attacks on AI Agent Systems (arxiv.org, 2026-10-12T04:00:00)
    Score: 10.48
    arXiv:2610.12524v1 Announce Type: new
    Abstract: As autonomous AI agents become more integrated into sensitive domains, it is crucial that single and multi-agent AI systems are secured. This paper investigates how prompt injection attacks can exploit both direct user inputs and inter?agent communication to leak sensitive information and trigger cross-agent infections. While our tests mostly highlight key vulnerabilities in single-agent systems, we also extend these attacks to multi-agent framewo
18. Introducing Claude Haiku 5.5 on AWS (aws.amazon.com, 2026-10-07T18:52:10)
    Score: 10.337
    Claude Haiku 5.5 is now available on Amazon Bedrock and Claude Platform on AWS. According to Anthropic, it is the fastest, most efficient model in the Claude 5.5 family, built for subagents and high-volume, cost-sensitive work, and costs around 75% less than Claude Haiku 4.5 for most tasks. This post covers its improvements and how to get started.
19. Introducing GLM 5.3 on Amazon Bedrock (aws.amazon.com, 2026-10-05T23:25:17)
    Score: 9.906
    GLM 5.3 from Z.ai is now available on Amazon Bedrock: a 753B-parameter mixture-of-experts model built for coding and long-horizon agentic tasks. Learn how to invoke it with the OpenAI-compatible APIs, cut cost and latency with prompt caching, and run an authorized security test with the open-source Strix agent.
20. Supercharge regulated workloads with Claude Code and Amazon Bedrock (aws.amazon.com, 2026-10-05T17:25:20)
    Score: 9.847
    Anthropic Claude Opus 5.5 and Claude Sonnet 5.5 are available on Amazon Bedrock in the AWS GovCloud (US) Regions. Learn how to use them with Claude Code, Anthropic's agentic coding tool, for compliance-aligned, AI-assisted development on regulated and ITAR workloads.
21. Enabling Extended SGX Enclaves inside Confidential Virtual Machines (arxiv.org, 2026-10-12T04:00:00)
    Score: 9.48
    arXiv:2610.12695v1 Announce Type: new
    Abstract: Confidential Virtual Machines (CVMs), such as AMD SEV-SNP, and process enclaves, such as Intel SGX, provide workload protection against cloud threats. However, they strongly differ in TCB, deployment cost and speed, peer workload protection, resource sharing, fine-grained isolation and attestation, and programming capabilities. Nesting process enclaves within CVMs offers an optimal solution and an opportunity to address long-standing SGX limitatio
22. Post-Quantum Session Establishment for Matter over Thread: Message Size, Transport, and Protocol Options (arxiv.org, 2026-10-12T04:00:00)
    Score: 9.48
    arXiv:2610.13026v1 Announce Type: new
    Abstract: Matter's first post-quantum change protects device attestation with ML-DSA, but its operational handshake, CASE, still derives session keys from P-256 ECDH, so recorded traffic can be decrypted once a quantum computer exists. We quantify the cost of moving CASE to ML-KEM and post-quantum signatures over Thread, where frames carry at most 127 bytes and a Matter message at most 1178 bytes. A byte-exact model of the CASE messages shows that a dr
23. The Limits of Query Monitoring against Distributed Model Extraction (arxiv.org, 2026-10-12T04:00:00)
    Score: 9.48
    arXiv:2610.13083v1 Announce Type: new
    Abstract: Model providers now face distillation campaigns run through thousands of accounts, each of which looks like an ordinary user. When network and payment metadata cannot link those accounts, the only evidence left is what they ask. We study what a provider can and cannot learn from those queries alone, using one million real chat turns for development and 17 later weeks that no design choice touched for a single frozen evaluation. The answer is set l
24. APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces (arxiv.org, 2026-10-12T04:00:00)
    Score: 9.48
    arXiv:2610.13623v1 Announce Type: new
    Abstract: Professional agents increasingly work through command-line interfaces and multiple tools over shared workspaces containing files, messages, and business applications. These environments create safety risks that are difficult to capture with attacks placed at a single, known read location. We introduce APEX-Harm, a benchmark for evaluating prompt and script injection against long-horizon agents performing finance, consulting, and legal tasks in lar
25. VEPA: Verified Evolving Formal Policy for Lifelong LLM Agent Guardrails (arxiv.org, 2026-10-12T04:00:00)
    Score: 9.48
    arXiv:2610.12813v1 Announce Type: cross
    Abstract: LLM agents increasingly operate over sensitive resources, such as private documents and mutable system states, which can create serious safety and security risks. Runtime guardrails should therefore make inspectable, reproducible, and deterministic allow or refuse decisions before potentially irreversible agent actions are executed. However, safety and security requirements are rarely complete before deployment, so guardrails need to keep adapti

Auto-generated 2026-10-12