🔍 Read the full analysis: It’s Worse Than We Thought: AI Has Taken The Clock Off Cyberattacks. What Companies And Users Should Do Now on ThorstenMeyerAI.com
Get privacy and security gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
US agencies, Microsoft and security researchers have reported cases in which AI-generated scripts or agent frameworks supported cyber operations. The accounts show attackers using automation for reconnaissance, login attempts and exploitation, but also document failed attempts and do not establish that AI makes attacks universally successful. The reports raise pressure on organizations to maintain basic security controls and respond quickly to vulnerabilities.
In an August 18 joint advisory, the NSA, CISA, FBI, Department of Energy and Environmental Protection Agency said actors were conducting reconnaissance and developing capabilities against internet-reachable Siemens S7 programmable logic controllers. The agencies said attackers used AI-generated exploitation scripts disguised as monitoring tools and built with the open-source snap7 library. They assessed the activity as likely preparation to cause operational effects, not proof that such effects had occurred.
Microsoft Threat Intelligence reported on October 21 that Storm-0940, which Microsoft describes as a China-state-aligned actor active since 2021, used the ARTEX open-source agentic penetration-testing framework in activity connected to a ConnectWise Cloud Backup compromise. Microsoft said the actor attempted to sign in to more than 2,250 tenants after obtaining credentials for the backup application. During a password-spray campaign from October 10 to 14, Microsoft recorded more than 151,000 failed logons, with a peak of 2,900 per minute; Microsoft said none of those attempts succeeded.
Microsoft said that in several attacks the time from initial targeting to privileged cloud access was under three hours. It also cautioned that most targeting was unsuccessful and that an affected application or successful sign-in does not by itself show that data was accessed. ConnectWise said existing backup archives showed no indication of compromise and said it suspended Cloud Backup, SaaS Security and Cloud Management while investigating.
Infographic 01 · The record so far
AI is in the attacker’s toolchain. On the record.
Six documented events in twelve weeks. Each label says what the source itself states about AI. Figures are the vendors’ and agencies’ own.
- 30 Jul 2026Unit 42: AI on attackDeepSeek inside the Hermes Agent framework picks and switches targets on its own. Limited impact: across autonomous and manual attempts, data theft from 3 NetScaler systems confirmed. Single-vendor report.
- 18 Aug 2026NSA, CISA, FBI, DOE, EPA: AI on attackAI-generated exploitation scripts, disguised as monitoring tools, against Siemens S7 PLCs in US critical infrastructure. “Not a theoretical risk.”
- 10–14 OctStorm-0940 password spray: AI on attack151,000+ failed logons across eight organizations. No successful sign-in. Lockouts made up nearly 98% of responses.
- 16–21 OctConnectWise Cloud Backup: AI on attackStolen app credentials, then machine-speed sign-ins to customer tenants. Microsoft attributes it to the China-state aligned Storm-0940.
- 21 Oct 2026OpenAI, Patch the Planet: AI on defenceAI found the critical PHP flaws before attackers did. OpenAI expects similar findings to “increase sharply in the coming months.”
- 21–22 OctQuest Diagnostics: no AI link reportedUnauthorized access to a lab ordering and results system; certain patient records accessed. No cause or attacker named.
Most AI-driven attempts in these reports failed. What changed is the cost of trying.Assessment: Thorsten Meyer
Sources: Unit 42 (30 Jul 2026); NSA, CISA, FBI, DOE, EPA advisory (18 Aug 2026); Microsoft Threat Intelligence (21 Oct 2026); ConnectWise Trust Center (from 16 Oct 2026); OpenAI, Patch the Planet (21 Oct 2026); Quest Diagnostics Form 8-K (22 Oct 2026).
Powered by Thorsten Meyer AI
Automation Compresses Attack Timelines
The reports matter because they describe AI as a tool for performing or speeding up ordinary attack steps, rather than as evidence of a wholly new kind of cyber operation. Scripts and agents can assist with reconnaissance, adapting exploits or repeated login attempts. Microsoft’s account of access in under three hours shows why a shorter response window can matter when an attacker reaches a privileged account.
That does not mean every AI-assisted attempt works. In the Microsoft-reported password-spray campaign, all logged attempts failed, and the company said most targeting was unsuccessful. For organizations, the practical concern is that automation can increase the pace or scale of attempts, making established defenses—asset inventories, timely patching, strong authentication and monitoring—more consequential.
Industrial systems bring additional consequences: a compromised controller could affect physical processes. The agencies identified manufacturing, energy, water, chemical, food and agriculture, and commercial facilities as sectors of concern. Their advisory describes suspected preparation and recommends protective measures; it does not report that attackers caused operational disruption.
business firewall appliance with intrusion prevention
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Three Reports, Different Evidence
The developments are not one coordinated incident. They are separate reports covering industrial control systems, a cloud backup supply-chain compromise and agent-assisted activity. The US agencies’ August advisory concerns Siemens S7 controllers, while Microsoft’s October report concerns Storm-0940 and ARTEX. Each source describes its own observations and assessments.
A further example comes from Palo Alto Networks’ Unit 42, which published a report on July 30. The source material says Unit 42 documented a Chinese-speaking actor using the DeepSeek model with the Hermes Agent framework, directed through Telegram. The account says the agent searched GitHub after an initial exploit attempt failed. The supplied material cuts off before giving the rest of that case, so no further outcome can be established here.
OpenAI said that AI had helped identify critical PHP vulnerabilities patched on October 21 and forecast that AI-discovered critical vulnerabilities would “increase sharply in the coming months.” That is a company forecast, not a measured prediction of how many flaws will be found or exploited. Likewise, a security leader’s expectation of a coming “vulnerability tsunami” and an approximately 18-month period of pressure is an expert opinion reported in the source material, not a verified timeline.
““This is not a theoretical risk—it is an active threat.””
— The five US agencies in the August 18 joint advisory
enterprise endpoint detection and response software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What The Reports Do Not Establish
The reports do not establish that AI has made cyberattacks universally faster, more successful or impossible to contain. Microsoft said most targeting failed, and the 151,000-plus password-spray attempts it logged did not succeed. Its report also says a successful sign-in or affected app does not necessarily mean data was accessed. ConnectWise reported no indication of compromise in existing backup archives, while its investigation was ongoing.
For the industrial activity, the agencies described reconnaissance and capability development and assessed likely preparation for possible operational effects. The supplied advisory details do not confirm damage to controllers or disruption of facilities. The available description of Unit 42’s case is incomplete, so its final outcome is also unclear. No evidence in the material supports treating the forecasts about future vulnerability discovery or an 18-month window as confirmed outcomes.
hardware security keys for Microsoft 365 multi-factor authentication
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Controls And Investigations To Watch
ConnectWise’s investigation and the status of the suspended services remain developments to watch; the source material gives no completion date or final findings. Organizations using affected services should follow vendor and Microsoft updates and review their own access logs, while distinguishing an attempted sign-in from confirmed data access.
The agencies advised operators to inventory controllers, patch firmware, keep programmable logic controllers off the public internet, separate IT and operational technology networks, require multifactor authentication for remote OT access and monitor S7 traffic for writes outside approved change windows. Microsoft’s account also makes third-party application access and privileged cloud permissions relevant areas for review.
For individuals, the source material recommends starting with the email account: use phishing-resistant sign-in where available, review recovery methods and connected applications, and respond promptly to security alerts. OpenAI’s forecast points to possible added patching demands, but organizations should base immediate decisions on vendor advisories and confirmed vulnerabilities rather than treating the forecast as a timetable.
industrial network security monitoring appliance for Siemens PLC
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Did AI cause a confirmed disruption to industrial facilities?
The August advisory describes reconnaissance and AI-generated scripts targeting Siemens S7 controllers, and assesses likely preparation for operational effects. The supplied material does not confirm a resulting disruption or physical damage.
What did Microsoft report about Storm-0940?
Microsoft said Storm-0940 used the ARTEX framework in activity tied to the ConnectWise Cloud Backup compromise and attempted access to more than 2,250 tenants. It also said most targeting was unsuccessful and that the recorded password-spray attempts all failed.
Does a successful sign-in mean data was stolen?
No. Microsoft cautioned that an affected application or successful sign-in does not necessarily mean data was accessed. The available account does not establish that data was taken in every case.
What should organizations do now?
Follow relevant vendor and agency advisories, inventory and patch systems, limit internet exposure of operational technology, review third-party application permissions, use strong multifactor authentication and practise incident response. The agencies specifically recommend monitoring industrial control traffic for unauthorized writes.
Is a surge in AI-discovered vulnerabilities confirmed?
No. OpenAI said it expects AI-found critical vulnerabilities to increase sharply, but that is a forecast. The source material does not provide a measured count or a confirmed timeline for such an increase.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
![Stalking The Wily Hacker: 40 Years Later – Cliff Stoll [Video]](https://cipherdot.com/wp-content/uploads/2026/08/stalking-the-wily-hacker-40-years-later-cliff-stoll-video-featured-560x300.jpg)
![216M Spy TVs – The LG Smart TV Problem [Video]](https://cipherdot.com/wp-content/uploads/2026/09/216m-spy-tvs-the-lg-smart-tv-problem-video-featured-560x300.jpg)
