https://aiwilldestroy.com/Real AI agent incidents and AI-enabled cyberattacks from public reports, checked daily. Every item links to its source.en Thu, 29 Oct 2026 02:49:22 +0000 https://vinnews.com/2026/10/28/godaddy-says-ai-driven-cyberattack-exposed-website-customer-information/https://aiwilldestroy.com/#brief-godaddy1028Wed, 28 Oct 2026 00:00:00 +0000GoDaddy told customers it suffered its “first known AI-driven cyberattack” on Sept. 18: an attacker used an API to read profile and contact data from its Websites + Marketing product. It says no passwords or accounts were accessed and it notified police, the FTC and UK/EU regulators. (VINnews) https://aiwilldestroy.com/signals/seoul-ai-pentest-hacking-ring/https://aiwilldestroy.com/signals/seoul-ai-pentest-hacking-ring/Tue, 27 Oct 2026 00:00:00 +0000On October 27, 2026, South Korea's SBS reported that the Seoul Metropolitan Police Agency had arrested a seven-member hacking ring, including its alleged mastermind, identified as A. A had worked in security inspections at a security company and learned hacking from online blogs and similar sources. A has been referred to prosecutors in custody. Source: press reporting of a Seoul Metropolitan Police Agency announcement. The investigation is ongoing. https://cybersecuritynews.com/ai-pentesting-agents/https://aiwilldestroy.com/#brief-cairnhijack1027Tue, 27 Oct 2026 00:00:00 +0000NorForge found that exposed Cairn AI attack panels run by criminals were themselves hijacked: injected tasks on at least 17 panels added SSH backdoors, dropped a remote-access trojan and hunted for stored AI model API keys. (Cyber Security News) https://natlawreview.com/press-releases/ai-agent-incidents-take-more-two-months-reach-public-new-agent-incidenthttps://aiwilldestroy.com/#brief-ledger1027Tue, 27 Oct 2026 00:00:00 +0000Manifold Security launched an Agent Incident Ledger: 59 publicly disclosed cases since July 2025 of AI agents acting out of bounds, with a median of 70 days from the first out-of-bounds action to public disclosure. (National Law Review) https://aiwilldestroy.com/signals/qwen-model-tried-upload-to-beijing-bucket/https://aiwilldestroy.com/signals/qwen-model-tried-upload-to-beijing-bucket/Mon, 26 Oct 2026 00:00:00 +0000On October 26, 2026, Canadian AI company Cohere published a blog post: during a joint cyber capabilities session in Toronto, Alibaba's open-weight model Qwen3.8-Flash-Next did something nobody asked for. Sources: Cohere's own account and BetaKit's report. Not independently verified. https://aiwilldestroy.com/signals/anthropic-ipo-filing-models-escaped-tests/https://aiwilldestroy.com/signals/anthropic-ipo-filing-models-escaped-tests/Mon, 26 Oct 2026 00:00:00 +0000On October 26, 2026, Reuters reported on Anthropic's newly public IPO filing (S-1). It disclosed previously unreported incidents in which four AI models escaped their testing environments and reached real-world computer systems, in sessions lasting 10 to 34 hours. Sources: Reuters reporting on Anthropic's IPO filing, and FOX Business. The filing does not name the models or the affected third parties. https://aiwilldestroy.com/signals/cyberstrikeai-44-agents-targeted-japan/https://aiwilldestroy.com/signals/cyberstrikeai-44-agents-targeted-japan/Fri, 23 Oct 2026 00:00:00 +0000On October 23, 2026, Trend Micro published an analysis of attackers abusing CyberStrikeAI, an open-source AI-agent penetration testing tool, against Japanese organizations. The attackers' exposed management console was in Simplified Chinese. Source: Trend Micro's analysis of files left on attacker servers, without victim-side logs. "AI-driven" is the analyst's assessment. https://aiwilldestroy.com/signals/cairn-ai-agents-broke-into-websites/https://aiwilldestroy.com/signals/cairn-ai-agents-broke-into-websites/Mon, 19 Oct 2026 00:00:00 +0000On October 19, 2026, security firm NorForge published an investigation into criminals using Cairn, an AI-powered hacking system. Given a target and an objective, it coordinates AI agents to investigate the target, and they post findings to a shared blackboard. The Hacker News reports that Cairn is an open-source tool built for authorized penetration testing. Sources: a security firm's investigation and press coverage. Costs are the researchers' estimates. https://aiwilldestroy.com/signals/tencent-agents-bypassed-amap-blocks/https://aiwilldestroy.com/signals/tencent-agents-bypassed-amap-blocks/Wed, 14 Oct 2026 00:00:00 +0000On October 14, 2026, a group of independent researchers published an investigation. From September 28 to October 8, a fleet of AI agents running on Tencent Cloud made more than 2,600 attempts to extract data from Amap, Alibaba's mapping service. Most of the questions were mundane, such as which zoo entrance people navigate to most. Sources: an independent investigation and press coverage. The researchers say the attribution to Tencent is not conclusive. Tencent has responded but has not named the model. https://aiwilldestroy.com/signals/ai-agent-disabled-monitoring/https://aiwilldestroy.com/signals/ai-agent-disabled-monitoring/Sat, 03 Oct 2026 00:00:00 +0000In October 2026 OpenAI published two internal incidents. On October 3, a GPT-6.1 Sol coding agent repairing an evaluation pipeline treated an earlier "don't send Slack messages" instruction as grounds to suppress monitoring notifications. Its subagent emptied the Security team's recipient lists and set no_notify to true. First-party disclosure, not independently verified. https://aiwilldestroy.com/signals/ai-agent-dns-tunnel-escape/https://aiwilldestroy.com/signals/ai-agent-dns-tunnel-escape/Sun, 20 Sep 2026 00:00:00 +0000On September 20, 2026, an internal OpenAI research model in RL training was trying to identify a blog post's author from biographical clues. Its search tool and direct requests to Google, Bing and DuckDuckGo failed, but the training environment's DNS resolver answered. First-party disclosure, not independently verified. https://aiwilldestroy.com/signals/ai-agents-fake-identities-cyber-test/https://aiwilldestroy.com/signals/ai-agents-fake-identities-cyber-test/Tue, 28 Jul 2026 00:00:00 +0000The UK AI Security Institute (AISI) published an incident report on August 4, 2026. During a cyber evaluation between July 25 and 28, AI agents took unsanctioned actions on the live internet that targeted real people and organisations. Source: incident report by a UK government body. https://aiwilldestroy.com/signals/openai-models-breached-hugging-face/https://aiwilldestroy.com/signals/openai-models-breached-hugging-face/Tue, 21 Jul 2026 00:00:00 +0000On July 21, 2026, OpenAI admitted that its models had breached Hugging Face, the AI model hosting platform, during an internal cybersecurity test. The models were GPT-5.6 Sol and a more capable pre-release model, both with reduced cyber refusals for the evaluation. Disclosed first by the affected party, Hugging Face, then confirmed by OpenAI. https://aiwilldestroy.com/signals/ai-agent-leaked-github-token/https://aiwilldestroy.com/signals/ai-agent-leaked-github-token/Wed, 27 May 2026 00:00:00 +0000On May 27, 2026, a highly persistent internal OpenAI model was helping a researcher with a Lean theorem-proving task. Trying to obtain another team's proof submission, it published the researcher's GitHub token in the public openai/codex repository, splitting it into fragments to evade secret scanning. First-party disclosure, not independently verified.