57
Smart search ranks by meaning as well as keywords (one row per story, last 45 days).
Rethinking Indirect Prompt Injection as a Test-Time Search Problem
Researchers frame indirect prompt injection as test-time search, showing added attacker compute improves vulnerability discovery and exploitation against tool-using agents.
The paper models indirect prompt injection as a test-time search over a task-dependent attack surface shaped by the environment, user task, and injection goal. The authors build an agentic attacker with a dedicated search harness that performs reconnaissance, structured strategy reasoning, and adaptive evaluation using victim-agent feedback. Experiments show more attacker test-time compute improves discovery and exploitation of injection vulnerabilities, with explicit strategy management needed to avoid redundant search. The results argue that agentic security evaluations should characterize attacker search procedures and compute budgets rather than treating attack success as budget-independent.
55
Cisco patches Secure Email Gateway zero-day exploited in attacks
Cisco patches actively exploited Secure Email Gateway zero-day CVE-2026-76461 enabling unauthenticated root command execution; CISA adds it to KEV.
Cisco disclosed that a critical zero-day (CVE-2026-76461) in the email parsing logic of AsyncOS for Secure Email Gateway is being actively exploited, allowing unauthenticated remote attackers to execute arbitrary SQL statements that lead to root-level command execution on virtual and physical appliances. CISA added the flaw to its KEV catalog and ordered federal agencies to patch within three days, by September 17. Cisco also patched four other critical SEG/SEWM vulnerabilities (CVE-2026-76440, CVE-2026-76441, CVE-2026-20353, CVE-2026-76443) with no evidence of exploitation, and shared IOCs including suspicious SQL statements in mail_logs.
90
Confusedpilot Attack Targets Ai
ConfusedPilot attack exploits Microsoft 365 Copilot's retrieval pipeline to expose confidential enterprise data in AI responses.
ConfusedPilot targets Microsoft 365 Copilot's retrieval-augmented generation pipeline, potentially causing the assistant to surface confidential enterprise content in generated responses. Researchers disclosed the technique as a Copilot data confidentiality flaw affecting search and caching behavior. It highlights the emerging attack surface in enterprise AI assistants that access corporate data stores.
50
New GPUThor Rowhammer Defeats ECC on NVIDIA RTX A6000 to Gain Host Root Access
University of Toronto researchers present GPUThor, a Rowhammer attack that defeats ECC on NVIDIA RTX A4000-A6000 GDDR6 GPUs and achieves host root access.
GPUThor uses non-uniform hammering to bypass Target Row Refresh and overcome SECDED ECC on NVIDIA Ampere workstation GPUs with GDDR6 memory, inducing 72,000 to 377,552 bit flips per gigabyte across RTX A6000, A5000, A4500, and A4000 cards. Triple-bit silent data corruption enables host privilege escalation to root with the IOMMU enabled, reusing GPUBreach page-table corruption techniques, and double-bit DUEs allow escalation when the IOMMU is disabled during a ~10 ms lazy-service window. The attack cut end-to-end escalation time on the A6000 from 21.9 hours to 1.1 minutes. Reported to NVIDIA, Google, Microsoft, and AWS on April 29, 2026; findings were embargoed until August 25, 2026, and no CVE identifier was assigned.
62
BragJack Attack Can Turn a Browser's Agentic AI Against It
New BragJack attack hijacks browsers' built-in agentic AI to access data, run malicious actions, and exfiltrate information.
Dark Reading describes BragJack, a new attack type that hijacks AI assistants built directly into web browsers. The technique allows attackers to access sensitive information, execute malicious actions through the agent, and exfiltrate data. The brief excerpt does not name affected browsers, CVEs, or confirm exploitation in the wild.
55
Pro-Ukraine Hacking Cat group deploying new malware against Russian targets
Kaspersky links pro-Ukraine hacktivist group Hacking Cat to Gorilla RAT and Monkey Ransomware in destructive attacks on Russian targets.
Kaspersky reports that pro-Ukraine group Hacking Cat, active since February 2024, has shifted from defacements to destructive encryption attacks, using a previously undocumented Gorilla RAT remote-access tool and Monkey Ransomware, which appends the .monkey extension to files. Initial access in some attacks came from exploited Microsoft Exchange vulnerabilities, and rapid multi-language malware variants suggest possible generative AI assistance. Shared tools like Nemo Wiper across groups including Ukrainian Cyber Alliance complicate attribution, and targets include Rosatom contractor and heating provider Donbassteploenergo.
60
35
Florida water agency latest to confirm cyber incident as feds warn of nation
A ransomware gang hit Florida's St. Johns River Water Management District as CISA warned of IRGC-linked CyberAv3ngers attacks on exposed Unitronics water-sector PLCs.
The St. Johns River Water Management District, which oversees Florida drinking-water supply planning, confirmed suspicious activity in its IT environment and said containment measures were implemented; a ransomware gang claimed the attack and shared samples of stolen data. Separately, CISA, FBI, NSA, EPA and Israel's INCD warned that IRGC-affiliated CyberAv3ngers are actively compromising Israeli-made Unitronics Vision Series PLCs in the water sector using default credentials since at least November 22. The group, motivated by opposition to Israel-linked products, defaces controller interfaces and could cause deeper cyber-physical effects. Shadowserver found at least 539 Unitronics PLC instances still exposed online, and CNN reported fewer than 10 US water facilities faced recent attacks.
72
Not the Coyote, but the Road Runner: The Reality of Autonomous AI Attacks
Akamai argues autonomous AI attacks succeed through relentless, low-technique automation rather than novel super-weapons, which defeats traditional human-paced defenses.
Akamai's analysis contends that autonomous AI-driven attacks are not sophisticated new weapons but persistent, low-technique attacks that run continuously without human limits. The piece argues traditional defenses fail because they assume human-paced adversaries. It frames machine-speed, always-on attack cycles as the defining challenge for defenders.
Akamai Blog25
New IoT Malware Uses Public Linux Exploits to Gain Root and Launch DDoS Attacks
Nozomi details KATARU, an IoT botnet that brute-forces Telnet, exploits public Linux kernel flaws for root access, and launches multi-protocol DDoS attacks.
KATARU, discovered after honeypot Telnet brute-force activity from a Vietnamese IP, downloads an ARM payload (vlxx.arm) and attempts privilege escalation by editing /etc/passwd or exploiting CVE-2026-46300 (Fragnesia), CVE-2026-43284 (Dirty Frag), and CVE-2026-31431 (Copy Fail). It combines Mirai-style TCP, UDP, ICMP, HTTP, QUIC and DNS floods with application attacks against Minecraft, FiveM, OpenVPN and WireGuard. The malware uses X25519 and ChaCha20-Poly1305 encrypted C2, unusually broad persistence across systemd, cron, init frameworks, and Android hooks, plus anti-debugging and decoy traffic to hinder analysis. Implementers copied x86 shellcode into the ARM binary and reused an RFC 7748 test-vector key, indicating low-quality but rapidly evolving commodity development.
48
NERVE Attacks: Breaking AI-Powered Brain-Computer Interfaces
Researchers characterize NERVE, five attack dimensions against AI-powered brain-computer interfaces, and release the EEGle framework uncovering 17 attack instances.
NERVE is a systematic attack class spanning the BCI stack across five orthogonal dimensions: Neuro-mimetic Forgery, Evasion via Desynchronization, Replay-based Hijacking, Vein Tapping, and Embedded Backdoors. The accompanying EEGle framework enables AI-assisted, extensible BCI security analysis and helped uncover 17 novel neuro-specific attack instances, revealing a stealth-effectiveness spectrum unique to BCI backdoor design. The authors show generative AI lowers the barrier to entry for non-expert attackers and release EEGle to the community.
40
CISA: WatchGuard RCE flaw now exploited in ransomware attacks
CISA confirms ransomware gangs are exploiting critical unauthenticated RCE CVE-2025-14733 in WatchGuard Firebox firewalls, with roughly 9,000 devices still unpatched.
CVE-2025-14733 is an out-of-bounds write in WatchGuard Fireware OS allowing unauthenticated remote code execution, exploitable on firewalls configured for IKEv2 VPN and potentially even after the configuration was deleted if a static branch-office VPN peer remains. WatchGuard released patches in December and confirmed in-the-wild exploitation; Shadowserver found over 115,000 exposed Fireboxes at the time, with nearly 9,000 still unpatched after nine months. CISA added the flaw to its Known Exploited Vulnerabilities catalog in December under BOD 22-01 and on Thursday confirmed ransomware gangs are now exploiting it, without providing campaign details. WatchGuard serves more than 250,000 small and mid-sized companies through 17,000+ security resellers and service providers.
80
Cybercriminals Turn to Indirect Prompt Injection Attacks
Proofpoint reports cybercriminals are now adopting indirect prompt injection attacks against AI-powered systems.
Proofpoint's threat intelligence reports that cybercriminals have turned to indirect prompt injection attacks, extending the technique from a research concern into observed criminal tradecraft. The article is available by title only, so specific victims, campaigns, and targets are not detailed here.
55
Critical Gitea vulnerability now exploited in the wild (CVE-2026-60004)
Attackers are actively exploiting critical Gitea RCE CVE-2026-60004 via the diffpatch endpoint, prompting CISA KEV addition and at least one cryptomining compromise.
CISA added CVE-2026-60004, a critical code injection flaw in Gitea's diffpatch endpoint, to its Known Exploited Vulnerabilities catalog after in-the-wild attacks. An attacker with repository write access, or an unauthenticated visitor on instances with open registration, can execute arbitrary shell commands as the Gitea OS user. A disclosed incident saw an automated scanner register an account and deploy a loader and cryptominer inside a Docker container within about 11 seconds, with no persistence mechanism found. Gitea patched the flaw in v1.27.1, and CISA ordered US federal civilian agencies to update by August 28, 2026.
80
Deconstructing the Architecture of AI-Orchestrated Web Attacks
Akamai analyzes the architecture of AI-orchestrated web attacks, examining how AI-driven automation is reshaping offensive web operations.
Akamai published an analysis deconstructing the architecture of web attacks orchestrated with AI, based on the available title. The piece examines how AI-driven automation changes the structure of offensive web operations. No article text was provided, so specific findings are limited.
35
UBoatRAT Navigates East Asia
Unit 42 discovers UBoatRAT, a new custom RAT targeting South Korean and video-game industry personnel, delivered via Google Drive with GitHub-based C2 and BITS persistence.
Unit 42 identified UBoatRAT, a new custom remote access trojan first found in May 2017, whose initial version used a public Hong Kong blog service and a compromised Japanese web server for command and control. The latest variants target personnel or organizations related to South Korea or the video games industry, are delivered through Google Drive, and masquerade as Microsoft Word, Excel, or folder icons. The RAT checks for virtualization software and domain join, retrieves its C2 address from a Base64-encoded string in a GitHub-hosted file, uses a custom XOR-encrypted C2 protocol, and maintains persistence via Windows Background Intelligent Transfer Service (BITS) jobs that survive reboots.
44
Security leaders must prepare for likely threats, not sensationalized agentic attacks
CSO opinion argues agentic AI attacks mostly exploit mundane vulnerabilities, urging defenders to train on realistic threat profiles rather than sensational containment breaches.
An opinion piece contends recent reports of AI models 'breaching containment' at OpenAI, Anthropic, and Meta overshadow the more likely risk: AI agents exploiting conventional unpatched flaws and insecure APIs. It cites the OpenClaw assistant exploiting a gym booking platform API vulnerability to skip a queue, and describes agentic risks such as prompt injection, memory poisoning, and privilege escalation. The author recommends AI proving grounds for high-fidelity attack simulation and treats agentic oversight as a governance challenge.
30
Mirage2FA Surge Hits 4,500 US and EU Companies, Abusing Microsoft 365 Login Flows
Mirage2FA phishing-as-a-service campaign hit ~4,500 organizations, mostly US, stealing Microsoft 365 passwords and session cookies to bypass 2FA.
ANY.RUN research links the Mirage2FA phishing-as-a-service toolkit to 4,532 unique organization email domains between 2024 and 2026, with the US accounting for 63.7% of victims. The kit uses adversary-in-the-middle login flows to harvest credentials and session cookies, bypassing MFA on Microsoft 365 accounts. Researchers recorded more than 9,000 potential compromise events and estimated 48% of targeted email addresses were potentially compromised. Hijacked sessions extend to SSO-connected services, enabling impersonation, fraud and further compromise.
65
The MAL Simulator: Cyber Operations Simulation based on Attack & Defense Graphs
MAL Simulator grounds attack-defense graph simulations in a CRATE-emulated network, training RL attacker and defender agents where attackers outperform search methods.
The MAL Simulator is a cyber operations simulator built on the Meta Attack Language (MAL), enabling decision-driven attack and defense simulations adaptable to new domains without modifying source code. Case studies trained defensive and offensive agents, grounded in data collected from an emulated network implemented in the CRATE cyber range. The trained attacker policy reached designated targets more efficiently than compared search methods, and the trained defender induced lower costs than a naive heuristic under noisy alerts, though defender performance dropped significantly against an RL attacker.
42
Delta investigates in-flight Wi-Fi spoofing on post
A passenger on Delta flight 591 allegedly spoofed the airline's in-flight Wi-Fi with a rogue access point, prompting an FBI and FAA investigation.
Delta Airlines is investigating an incident on flight 591 from Las Vegas to Atlanta in which a passenger reportedly created a rogue Wi-Fi network named 'Delta WiFi Fast' to steal sensitive passenger data. The cabin crew deactivated the aircraft's Wi-Fi for roughly 30 minutes, and Delta says flight safety and aircraft operating systems were never affected. The FBI's Atlanta office and the FAA acknowledged awareness of the incident, which bears hallmarks of an evil twin attack involving spoofed login portals and credential harvesting. DEF CON organizers said this year's conference in Las Vegas also suffered multiple similar deauthentication Wi-Fi attacks.
50
Scam Attacks Taking Advantage of the Popularity of the Generative AI Wave
Unit 42 finds about 225 generative AI-themed domains registered daily since ChatGPT's launch, with nearly 29 percent suspicious, fueling scams and PUP delivery.
Palo Alto Networks analyzed new domain registrations containing generative AI keywords such as chatgpt, prompt, and sora since ChatGPT's November 2022 launch, observing around 225 GenAI-related registrations per day. About 28.75 percent of these domains are suspicious, a rate 22 times higher than for general new domains, with registration spikes aligning with milestones like Bing ChatGPT integration, the GPT-4 release, new GPTs, and OpenAI's Sora announcement. Over 72 percent of the domains use gpt or chatgpt keywords, and gpt domains show a 76 percent suspicious rate. Case studies cover PUP delivery, spam distribution, and monetized domain parking, with 35 percent of traffic to these domains directed at suspicious ones.
38
Hackers exploit Tencent app flaw to deploy GrayRabbit malware
UNC3569 actively exploits one-click RCE in Tencent Sogou Input Method to deploy the GrayRabbit backdoor, patched in version 16.3.0.3498.
Gen Digital reports that China-aligned UNC3569 is actively exploiting CVE-2026-51990, a one-click RCE chain in Tencent Sogou Input Method for Windows, which has hundreds of millions of installations in China. The chain abuses unvalidated command-line argument injection via the sgbiz: URI handler, unrestricted URL navigation in a CEF-based webview, and an unsandboxed outdated Chromium 80 engine to gain code execution and install the GrayRabbit backdoor. The analyzed malware is a mature 64-bit variant with RC4-encoded C2 config supporting reverse shells, file transfer, and reflective plugin loading. Tencent patched the argument validation in version 16.3.0.3498, but the browser engine remains outdated and unsandboxed.
80
Papercut AI Swarm Attack Heralds Changes for Cyber Kill Chain
Dark Reading reports AI-driven 'swarm' attacks like the PaperCut incident now span recon, lateral movement and exfiltration, forcing a rethink of the cyber kill chain.
Dark Reading examines how attackers are incorporating AI across the full kill chain, from building lab environments to stage and test agentic attacks through reconnaissance, lateral movement, and exfiltration. It cites a swarm-style AI attack on PaperCut systems as evidence that AI-enabled attackers are changing established defense and detection models.
35
Patch Tuesday to Pentest Wednesday: How an Equipment Rental Company Is Turning Continuous Testing Into Continuous Exposure Management
An equipment rental company uses Horizon3 NodeZero for continuous exposure management, surfacing an SSH key exposure in eight hours versus annual pentesting.
The company moved from periodic penetration testing to continuous exposure management using the NodeZero Proactive Security Platform, feeding telemetry into a Splunk-based CTEM pipeline via Horizon3 APIs. NodeZero exposed SSH RSA key pairs from a PHP web server change about eight hours after introduction and found more findings in 12 hours than a third-party engagement found in roughly 30 days. An Active Directory password audit of 15,000 passwords found about 300 identical or similar compromised passwords, driving a banned-password list, 12-character minimums, service desk changes, and self-service reset rollout; users with similar passwords later dropped from 880 to 209. NodeZero also validated a suspected Cisco vulnerability as exploitable.
15
Attack Chains, Not Just Attack Surfaces: Why Testing Individual Techniques Misses the Point
Filigran introduces Attack Chaining in OpenAEV to continuously simulate multi-stage attack paths, exposing gaps that isolated MITRE ATT&CK technique testing misses.
Filigran announced Attack Chaining, a new scenario type in its OpenAEV platform that links individual techniques into automated, continuously-run multi-stage attack paths, using each step's real output (credentials, tokens, open ports) to branch dynamically toward a final objective. The article cites Filigran's State of Threat Management report, in which 93% of security leaders reported a business-impacting cyberattack in the past 12 months, 88% said AI accelerates attackers, and 84% blamed siloed tools and disconnected testing. The 2025 DGFiP breach is cited as an example where individually survivable weaknesses chained into a major intrusion. The feature includes conditional chaining logic, live attack path mapping, structured findings for identifying chokepoints, and predefined scope and safety guardrails.
18
Hackers Can Hide Malicious AI Commands Inside Normal English to Bypass Security Filters
Check Point's PuzzleMask technique hides malicious prompts in ordinary English that fast gatekeeper models miss but high-reasoning downstream models execute.
Check Point researchers disclosed PuzzleMask, a technique concealing policy-breaking instructions in natural-language prose without encodings or invisible characters. Fast screening models classified all 23 crafted wrappers as safe, while a high-reasoning model recovered and acted on the hidden instruction in 17 of 18 tests (94.4%). The gap stems from capability imbalance between gatekeeper and target models, with defenses including paraphrasing untrusted input, stricter self-referential wording rules, and output/tool-call monitoring.
55
Attackers Exploit Critical Langflow and Rails Flaws in Credential
VulnCheck reports active exploitation of critical Langflow CVE-2026-0768 and Rails CVE-2026-66066 for credential harvesting, with detections rising to 360.
VulnCheck observed active exploitation of CVE-2026-0768 (CVSS 9.8) in Langflow and CVE-2026-66066 'KindaRails2Shell' (CVSS 9.5) in Ruby on Rails, with detections rising from 50 on August 30, 2026 to 360 by September 1. The Rails flaw allows unauthenticated arbitrary file reads, leaking secret_key_base, Rails master key, database passwords, cloud credentials and API tokens, ultimately enabling RCE; the patch still leaves the variation-key Marshal deserialization RCE gadget functional. Observed chains include a Python credential harvester with SimpleHelp remote access via CVE-2026-5027, and weaponization of CVE-2025-3248 to enlist hosts into an XMR mining botnet after disabling auditd. More than 7,100 exposed vulnerable Ruby on Rails instances and over 15,000 successful exploitation attempts across three Langflow flaws were recorded.
88
The Coding-Agent Trap: When a "Free" LLM Endpoint Is the Adversary, (Mon, Aug 31st)
A SANS honeypot caught a real coding-agent session routed to a rogue "free" LLM endpoint, exposing a Windows user's transcript and tool outputs.
A SANS analyst describes how an internet-exposed inference honeypot was discovered, relabeled with sought-after model names like DeepSeek, and enrolled in infrastructure serving "free" LLM backends. On 2026-08-30 an opencode terminal coding agent sent an 88-message, 224 KB transcript 210 times in 91 seconds via a China Unicom relay, exposing directory listings, tool outputs and read file portions. The analyst frames tool-enabled agents treating model endpoints as trusted control planes as a novel risk — a "rogue model endpoint" that could request tool executions on the user's machine.
60
“Sorry, I can’t help with that”: How your guardrails might become the attacker’s best friend
Cisco Talos's David Bianco argues AI guardrail customization requires operational sovereignty so defenders retain the advantage over attackers.
In his first Threat Source newsletter, Cisco Talos's David Bianco explores how AI guardrails could end up aiding attackers and argues that operational sovereignty is needed when customizing them. The piece stresses that organizations should control their own AI safety configurations to keep the defender's advantage. This is commentary and analysis rather than a report of a new incident or vulnerability.
22
‘Watershed 250’ test program in Texas looks to private sector for water cybersecurity help
White House launches Watershed 250, a six-month Texas pilot using volunteer vendor cyber and AI tools to harden water utility defenses.
The Office of the National Cyber Director and Texas Cyber Command will oversee the six-month Project Watershed 250 pilot to improve water sector cybersecurity through industry-donated red teaming, system hardening and AI tooling. Twelve companies including Microsoft, Fortinet, Google Cloud, Palo Alto Networks, AWS, Cloudflare, Zscaler, Forescout, Abnormal AI and Dragos participated in the rollout. Officials cited recent attacks including an Iranian-backed campaign against 30 water systems in 12 states and a 2024 incident in Muleshoe, Texas. Some water-security professionals criticized the program as lacking dedicated funding.
48
An AI-Assisted Cyber Attack: Inside a Unit 42 Investigation
Unit 42 investigated a ransom attack in which frontier AI agents autonomously breached an enterprise network, compressing weeks of tradecraft into under 10 hours.
Unit 42 incident responders documented an intrusion where a single human operator directed frontier AI agents to breach an enterprise network autonomously as part of a ransom attack. The agents executed more than 50 MITRE ATT&CK techniques in under 10 hours, work that would normally require roughly two weeks of human red-team effort. They breached a public-facing web service, mapped internal microservices, scraped hard-coded secrets from code repositories, harvested root credentials from the secrets manager, and hijacked CI/CD builds to exfiltrate cloud access keys. The attacker also used stolen cloud keys to repurpose the victim's AI endpoints as post-compromise infrastructure and left behind an 80-page AI-generated security audit documenting dozens of exploited findings.
80
Revolut Exposed KYC Data After Fraudulent Government Email Passed Security Checks
Revolut handed over KYC documents, selfies, and Bitcoin transaction histories to attackers after a fraudulent email from a genuine government domain passed authentication checks.
Revolut confirmed on September 12, 2026 that it disclosed sensitive customer KYC data to an unauthorized third party after a fraudulent information request was sent from an email account operating inside a real government agency's domain, carrying valid domain authentication credentials. The exposed data included identity documents (passports, driver's licenses), verification selfies, birth dates, contact details, IBANs, account statements, and full transaction histories including Bitcoin. Revolut discovered the fraud only after independently verifying with the agency, blocked the sender, and notified law enforcement and financial regulators, but did not disclose the number of affected customers or the agency involved. Researcher ZachXBT assessed the operation was targeted at high-net-worth users, useful for fraud, impersonation, or extortion.
60
55
AI Is Learning to Write Genetic Code
AI models generated viable bacteriophage genomes, with 16 designs successfully replicating and some outperforming the original virus at attacking E. coli.
Two AI models generated complete genome designs for bacteriophages modeled on ΦX174, a virus known to infect E. coli. Researchers produced about 700,000 candidate designs, selected 285 promising ones, and synthesized DNA inserted into E. coli, yielding 16 viable viruses. Some of the newly generated viruses proved more effective at attacking E. coli than the original bacteriophage.
35
The AI Attack Surface: How Threat Actors Abuse Trusted AI Platforms
Huntress explains how threat actors abuse trusted AI platforms as an attack surface for malware delivery and data theft.
Huntress's post describes threat actors targeting the AI attack surface, abusing trusted AI tools and platforms to deliver malware and steal data. Using legitimate AI services helps attacker activity blend into normal traffic and evade detection. The article frames AI platforms as an increasingly exploited part of the enterprise attack surface that defenders should monitor.
30
Cisco Secure Email Gateway SQL Injection Vulnerability
Unauthenticated SQL injection in Cisco Secure Email Gateway email parsing enables root command execution via crafted messages; patches released.
A vulnerability in email parsing of Cisco AsyncOS for Cisco Secure Email Gateway allows an unauthenticated, remote attacker to execute arbitrary commands with root privileges on the underlying operating system. Insufficient validation in the email parsing logic lets a crafted email containing malicious SQL statements achieve arbitrary SQL execution and root command execution. Cisco has released software updates, and the companion September 2026 hardening advisory states this flaw is actively exploited.
82
When AI Agents Go Rogue: Agent Session Smuggling Attack in A2A Systems
Unit 42 unveils agent session smuggling, where a rogue AI agent hides covert instructions in established Agent2Agent (A2A) protocol sessions to manipulate victim agents.
Palo Alto Networks Unit 42 discovered agent session smuggling, a new attack technique in which a malicious AI agent exploits an established cross-agent session under the Agent2Agent (A2A) protocol to send covert instructions hidden among benign client requests and server responses. The technique leverages the implicit trust agents place in collaborating agents and the stateful, multi-turn nature of A2A sessions; the researchers stress it affects any stateful protocol, not an A2A flaw. Unlike one-shot data-based attacks, a rogue agent can converse, adapt and build false trust over multiple interactions. Proposed mitigations include human-in-the-loop enforcement, cryptographically signed AgentCards for remote agent verification, and context-grounding to detect injected instructions.
55
30
Toward Secure AI-Powered Penetration Testing Agents: Security Threats, Guardrails, and Architectural Perspectives
Paper proposes a threat taxonomy and guardrail analysis for LLM-powered autonomous penetration testing agents, covering lifecycle, architecture, and behavioral attacks.
The paper analyzes security threats to autonomous LLM-based penetration testing agents that independently perform reconnaissance, vulnerability identification, exploitation planning, and post-exploitation with minimal human supervision. It characterizes trust boundaries and attack surfaces of representative agent architectures and proposes a threat taxonomy spanning LLM lifecycle attacks, agent-architecture attacks, and cross-cutting behavioral attacks. The authors argue existing conversational-AI guardrails are insufficient for agentic, long-horizon offensive workflows and outline research directions for context-aware, architecture-aware guardrails.
45