AI in real attacks

Week of Nov 95

• Nov 11AttacksArctic Wolf ties hosts serving AntiFox, an open-source AI penetration-testing platform, to intrusions that in some instances reached a production AWS secret
• Nov 10AttacksRailway says the attacker in its October breach used abliterated models and an ARTEX-like harness, and puts unauthorised AI spend at about $740,000
• Nov 9AttacksGerman web host Mittwald says an attacker came in through the MCP server connecting customers' AI assistants to its platform and compromised 56 customer projects
• Nov 9AttacksElastic says infostealer operators logged in an average of eight minutes after infection, and in two cases worked through Claude Code and the ChatGPT desktop app on victims' machines
• Nov 9MarketsKorea's largest banks plan sharply larger 2027 information-security budgets after the financial-sector hacking wave

Week of Nov 217

• Nov 8AttacksMalbear Labs says a jailbroken DeepSeek Flash agent ran more than two weeks of largely unsupervised intrusion against a European telecom provider and organisations it hosts
• Nov 8AttacksVulnCheck's decoys caught an IP address from Microsoft's agentic secret-hunting report running an 11-stage Langflow exploitation session
• Nov 6AttacksWorkOS says the agent behind its breach belonged to a security researcher running benchmark evaluations with ARTEX, and had escaped its isolated test environment
• Nov 6AttacksChainalysis now says password hashes and live tokens for all its users were exposed, and requires users to reset passwords and rotate API keys
• Nov 6AttacksMicrosoft says a Cambodia-based criminal cluster, Storm-3358, began using ARTEX against financial organisations in October
• Nov 6AttacksCensys finds more than 800 ARTEX deployments reachable from the web without a password (via TechCrunch)
• Nov 5PolicySouth Korea's financial regulator orders security inspections of about 580 firms after 40 breach reports at 37 financial companies since September 30
• Nov 5AttacksMicrosoft says attackers have used an agentic framework since August to mine public GitHub for secrets, in one case reaching Azure Key Vault data in five minutes
• Nov 5AttacksCheck Point finds 1,608 live ARTEX panels, and says one operator switched to more permissive Chinese models when Anthropic's Opus 5.5 hit guardrails
• Nov 5AttacksA widely forked ARTEX copy pulled a backdoored Docker image giving a root shell on every install, and 62 exposed consoles were running it
• Nov 4AttacksChainalysis says an "on-path attack" exposed some customer passwords, and takes KYT and its login page offline over abnormal activity
• Nov 4AttacksAcronis says one operator in Taiwan ran a mesh of 12 AI agents against hundreds of organisations, reaching Taipei emergency-service records
• Nov 4AttacksIndia's CERT-In says an AI-assisted actor it calls CRAFTY FALCON is still attacking financial firms through their own payment APIs
• Nov 3AttacksGitBook says a "highly targeted attack using AI agents" may have exposed customer API tokens and integration secrets
• Nov 3PolicyThe UK NCSC judges AI-powered malware highly likely to be used in a ransomware attack on UK critical infrastructure by 2028, as the government funds a Cyber Shield and a vulnerability centre
• Nov 3AttacksBitdefender says a 3AM ransomware intrusion at a UK school estate behaved like an agentic loop
• Nov 2AttacksKaspersky says attackers used the CyberStrikeAI framework against schools and universities in Latin America and Africa

Week of Oct 268

• Nov 1PolicySouth Korea's financial regulators call the heads of the five largest banks to an AI-hacking response meeting on November 5
• Oct 30AttacksA ransomware operator's exposed workspace shows an AI coding agent wired to 154 models and more than $7,216 in AI spend
• Oct 30AttacksChainalysis says an attacker using "offensive agentic techniques" got into its systems and internal tools
• Oct 29AttacksCrowdStrike names a China-nexus actor, VAPOR PANDA, running the ARTEX agent framework against more than 200 organisations in 11 days
• Oct 29AttacksCisco Talos describes NIK C2, a command-and-control framework that routes work through OpenAI's Chat Completions API
• Oct 27AttacksAn open-source attack tool using 44 AI agents was run against at least 19 organisations, most in Japan, TrendAI says
• Oct 26AttacksSeoul police allege a ring used AI tools to hack 60 of more than 90 targets and took 8.2 million records
• Oct 26AttacksDream says a suspected China-nexus actor chained the PHP2Shell flaws with a kernel bug to root 25 PAN-OS firewall and management servers, with Claude Code in its attack infrastructure

Week of Oct 198

• Oct 25AttacksAttackers used agentic AI platforms after breaking into a European government entity through the PHP flaws, Unit 42 reports
• Oct 23PolicyThe SEC's 2027 examination priorities say examiners will test whether firms are ready for AI-powered cyberattacks
• Oct 21AttacksMicrosoft says a China-state aligned actor automated attacks on hundreds of organisations with the ARTEX agent framework, through ConnectWise
• Oct 21AttacksSymantec says it captured an attack tool built to run up to 100,000 actions without human approval, paid for with other people's model endpoints
• Oct 21AttacksRapid7's quarterly threat report says AI was an operational component in 8% of the incidents it tracked
• Oct 20AttacksOne operator ran AI agents against more than 1,200 shops for about $5,000 in tokens, NorForge and Reuters report
• Oct 20AttacksRansomware affiliates left their AI sessions exposed: 8,257 messages, commercial models through OpenRouter and over $23,920 of spend
• Oct 19AttacksA Japanese security firm says it found ARTEX on infrastructure used against about twenty Japanese companies

Week of Oct 128

• Oct 16AttacksThe FBI and CISA tell election officials that “super intelligence” is making the phishing aimed at them more convincing
• Oct 15AttacksBitsight finds an autonomous pentest platform sold on a closed forum that runs on the buyer's stolen OpenAI, Anthropic and Google keys
• Oct 14DefenseMicrosoft disrupts a phishing marketplace it ties to a billion messages, and says the fake tenants behind it were likely stood up with AI-generated content
• Oct 14AttacksA survey of ransomware victims reports AI on the attacker's side in 84% of cases, and the victim's own AI exploited in 44%
• Oct 14AttacksZscaler puts a figure on the ARTEX campaign against Korean financial firms: personal data on about 67,000 people
• Oct 13AttacksThe Justice Department charges an OpenAI engineer over a recorder hidden in an executive conference room, and alleges a linked chatbot account was prompted on building and obfuscating it
• Oct 13DefenseATT&CK v20 splits prompt injection into two techniques and adds a campaign for an AI-assisted breach that reached a municipal water utility
• Oct 12AttacksA Russian-speaking access broker ran a 51-agent Claude Code cluster and took more than 628,000 records on Ukrainians, SOCRadar reports

Week of Oct 57

• Oct 8CapabilityARTEX's developer takes the tool closed-source and pulls it from GitHub, and an expert notes that copies already downloaded cannot be recalled
• Oct 7AttacksKorean investigators name ARTEX, a Chinese-language autonomous penetration-testing tool, in the run of bank breaches (via The Korea Times)
• Oct 7AttacksCrowdStrike says the operator behind the South Korean bank intrusions ran ARTEX on DeepSeek and worked through Claude Code
• Oct 6MarketsA rating agency says the AI-linked Korean bank breaches have produced no material losses and are unlikely to move the banks' ratings
• Oct 6PolicySouth Korea's prime minister orders an overhaul of the national information-protection system over the financial-sector attacks (via The Korea Times)
• Oct 5AttacksWikimedia says OpenAI agents edited its wikis, tried to turn its public pad into a proxy, and may have helped take a query service down
• Oct 5AttacksSouth Korea's financial regulator meets on a run of bank breaches where an AI penetration-testing tool is suspected and unconfirmed

Week of Sep 288

• Oct 1AttacksA DFIR firm reconstructs how rogue agents got onto the open web, and says public records cannot show whether they meant to cover their tracks
• Oct 1CapabilityMicrosoft says the threat landscape has moved toward autonomous execution and weaponisation now takes under a day
• Sep 30AttacksDIVD names the two Zammad zero-days behind the agent-run intrusion against it
• Sep 30AttacksAgents probed Canada's national archive, and Ottawa has opened a review (via The Next Web)
• Sep 30AttacksTransluce counts more than 200,000 agent requests to a US Education Department site, a SQL injection probe among them
• Sep 29AttacksA Dutch vulnerability-disclosure nonprofit says an autonomous agent, not a person, ran the attack on it
• Sep 28AttacksA malicious Custom GPT hosted on chatgpt.com walked users into installing a remote access trojan
• Sep 28AttacksAn Android banking trojan calls Gemini from the infected phone, while its operator console uses a model to rank which victims are worth attacking

Week of Sep 216

• Sep 25AttacksMicrosoft tracks the first documented agentic ransomware crew into an Azure tenant's service principals
• Sep 23AttacksA Docker botnet installs an off-the-shelf AI agent and tells it to take AI API keys first
• Sep 22AttacksCisco Talos documents a Windows implant that lets four AI models vote on its next move
• Sep 22DefenseTalos open-sources a framework for hunting AI-integrated malware
• Sep 22DefenseMicrosoft disrupts EvilTokens, a phishing service that sold an AI assistant with the kit
• Sep 22AttacksOne operator ran three open-source AI harnesses against online retailers at about $25 a company

Week of Sep 142

• Sep 15AttacksResearchers find an uncensored AI service sold by subscription on a criminal forum as an alternative to jailbreaking
• Sep 14AttacksSpain's data protection agency records its first notified personal-data breach executed by an AI agent

Week of Sep 77

• Sep 11AttacksA Brazilian firm documents a credential-stuffing platform that picks its targets and logs in with commercial models, testing 12.3 million credentials
• Sep 10AttacksAnthropic says a majority of the misuse operations it disrupted were executed or orchestrated by AI
• Sep 10AttacksAn espionage group ran its own autonomous vulnerability research program against a major security product
• Sep 9AttacksHundreds of AI agents drive a PaperCut campaign reaching 440 instances in 48 countries
• Sep 9PolicyA bipartisan House bill would have NIST write standards for finding, verifying and cutting off AI agents
• Sep 8AttacksGoogle records an attacker planning, building and running a mass credential-harvesting campaign with an autonomous multi-agent framework in under six hours
• Sep 8AttacksGoogle says a PRC-nexus actor runs open-weight models on victim compute to escape API monitoring, and that AI models and prompts are now extortion targets

Week of Aug 313

• Sep 3AttacksUnit 42 finds two criminal clusters in Latin America running intrusions with commercial chatbots
• Sep 2AttacksUnit 42 investigates an intrusion that ran more than 50 ATT&CK techniques in under ten hours
• Sep 2DefenseBooz Allen launches a counter-AI product and reports playbooks that cut autonomous-attacker success by more than 95%

Week of Aug 244

• Aug 27AttacksRansomware operators ran Cursor Agent inside victim networks to carry out hands-on intrusion steps
• Aug 27DefenseNIST says organisations are repeating decades-old identity mistakes with AI agents
• Aug 26AttacksFBI, NSA and Cyber National Mission Force say a China-linked group has been integrating AI into its operations
• Aug 25AttacksJoe Security analyses ToxNetV2, a Linux botnet that queries a jailbroken hosted LLM to propose attack commands

Week of Aug 173

• Aug 20AttacksCisco Talos finds a Chinese-speaking crew running agentic-AI tools in live post-compromise operations
• Aug 20DefenseUK NCSC issues interim guidance on securing agentic AI, including keeping the ability to “pull the plug”
• Aug 17AttacksRapid7 finds a crypto-fraud crew used Claude Code to build and run a vishing pipeline against wallet users

Week of Aug 103

• Aug 13AttacksIsraeli firm Dream reports China-linked operators ran a near-autonomous AI-agent intrusion of Taiwan's government
• Aug 13AttacksTrellix reports purpose-built offensive AI tools are being sold on criminal forums
• Aug 10AttacksA personal AI agent told only to book a gym class autonomously exploited the booking API to cancel another member's reservation

Week of Aug 31

• Aug 3AttacksCrowdStrike's 2026 Threat Hunting Report says AI is now embedded across adversary operations

Week of Jul 271

• Jul 30AttacksUnit 42 reports Chinese-speaking actor running autonomous attacks with DeepSeek and the Hermes Agent framework

Week of Jul 202

• Jul 24AttacksOpen-source Hermes agent run in "YOLO mode" automated an intrusion at Thailand's finance ministry
• Jul 21AttacksLLM-run agent deploys "ENCFORGE" ransomware built to encrypt AI/ML model stacks

Week of Jul 131

• Jul 14AttacksHunt.io reports suspected China-linked operators running Claude Code and DeepSeek as an intrusion toolchain against government targets in four countries

Week of Jun 291

• Jul 1AttacksSysdig documents JADEPUFFER, an LLM-driven agent that autonomously exploited Langflow and extorted a production database