🔍 Read the full analysis: The Faster Cyberattack Era: How AI Changes The Risks on ThorstenMeyerAI.com
Get smart everyday buys 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
A US interagency advisory says attackers are using AI-generated scripts against internet-reachable Siemens industrial controllers, and Microsoft reports that a China-state-aligned actor used an AI agent in a campaign involving hundreds of organizations. Microsoft says most targeting failed, while OpenAI reports that AI helped find critical software flaws before attackers did. The reports point to faster attack and defense workflows, but do not establish that AI makes attacks more likely to succeed in every case.
A joint advisory published August 18 by the NSA, CISA, FBI, Department of Energy and Environmental Protection Agency says threat actors used AI-generated scripts against internet-reachable Siemens S7 programmable logic controllers, or PLCs. These industrial computers can control equipment and processes at facilities such as water plants and factories. The advisory says actors scanned for exposed devices and used scripts based on the open-source snap7 library, disguised as monitoring tools. The agencies describe the activity as active and assess that it may be preparation for operational effects; that assessment is not confirmation of a successful disruption.
In an October 21 report, Microsoft Threat Intelligence linked China-state-aligned actor Storm-0940 to use of ARTEX, an open-source, multi-agent penetration-testing framework, in activity involving the ConnectWise Cloud Backup compromise. Microsoft says the actor attempted sign-ins to more than 2,250 tenants using credentials associated with the app. It recorded more than 151,000 failed logons from October 10 to 14, with a peak of 2,900 per minute; Microsoft says none of those attempts succeeded. In several cases, the time from initial targeting to privileged cloud access was under three hours.
Microsoft cautions that most of the targeting was unsuccessful and that an affected app or successful sign-in does not necessarily mean data was accessed. ConnectWise, which disclosed the compromise on October 16, said existing backup archives showed no indication of compromise. The company said it suspended Cloud Backup, SaaS Security and Cloud Management while investigating. Separately, OpenAI said AI helped find critical PHP vulnerabilities patched on October 21 before attackers did, and forecast that similar discoveries would rise sharply in coming months. That is a prediction, not a measured future trend.
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
Faster Workflows Raise the Stakes
The reports point to a change in speed and scale, rather than evidence of a new kind of attacker capable of bypassing every defense. AI tools can automate tasks such as reconnaissance, password attempts and adapting scripts. Microsoft says agents can work continuously; Palo Alto Networks’ Unit 42 reported that one agent completed hundreds of hours of manual targeting analysis in minutes. Those observations describe specific cases, not a universal measure of AI performance.
For organizations, faster attempts can compress the time available to identify exposed systems, revoke compromised credentials and respond to alerts. Industrial environments carry added operational risks: a compromised controller could affect physical processes, though the advisory does not report that these incidents caused such effects. At the same time, AI-assisted vulnerability discovery may bring more flaws to defenders’ attention and increase pressure to assess and install patches promptly.
The practical implication is not that every company needs an AI security product. The source material points instead to basic controls: accurate inventories, timely patching, stronger sign-in protection, careful review of third-party applications, separation between business and industrial networks, and rehearsed incident response. These measures address the routes attackers used, whether or not AI was involved.
industrial network security appliance OT ICS firewall
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Three Reported AI Use Cases
The developments span different parts of an attack chain. The August government advisory concerns industrial reconnaissance and scripts aimed at Siemens PLCs, including models in the S7-200 to S7-1500 range. It recommends keeping controllers off the public internet, patching firmware, blocking TCP port 102 at the perimeter, separating information-technology and operational-technology networks, and requiring multifactor authentication for remote access. It also advises monitoring S7 traffic for writes outside approved change windows.
Microsoft’s October report describes a separate route: abuse of trust already granted to a third-party cloud application. The company says ARTEX can support activity from reconnaissance through data exfiltration with limited human involvement. The ConnectWise investigation remains a distinct part of the account: Microsoft attributed the activity to Storm-0940, while ConnectWise reported no indication of compromise in existing backup archives and suspended services during its investigation.
A July 30 Unit 42 report provides another example, describing a Chinese-speaking actor who connected the DeepSeek model to the Hermes Agent framework and directed it through Telegram. Separately, OpenAI’s account highlights AI’s defensive use in vulnerability discovery. Together, the reports document uses across attack and defense; they do not establish that every AI-assisted operation is autonomous or successful.
““This is not a theoretical risk—it is an active threat.””
— The joint advisory by the NSA, CISA, FBI, DOE and EPA
Siemens S7 PLC industrial cybersecurity monitoring software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Impact and Forecasts Still Unsettled
The public reports do not establish that the PLC activity caused physical or operational damage, or that data was exfiltrated in the ConnectWise-related activity. Microsoft says most attempts failed and warns that a successful sign-in alone does not confirm access to data. ConnectWise’s statement about its backup archives is the company’s account; details of the investigation and its final findings were not provided in the supplied material.
It is also unclear how often AI tools are used across cyber operations, how much they improve success rates compared with conventional methods, or how many organizations were affected beyond the reported targeting. OpenAI’s expectation of a sharp rise in AI-found vulnerabilities is a forecast. A security leader’s prediction of a near-term “vulnerability tsunami” and roughly 18 months to prepare is an expert opinion, not a measured timeline or consensus forecast.
cloud security posture management software for Microsoft 365
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Investigations and Patching Continue
ConnectWise’s investigation will need to establish the scope and effects of the compromise, including whether any customer information was accessed. Microsoft’s report and the agencies’ advisory give organizations indicators and recommended controls, but the supplied material does not specify a date for a final government assessment or a complete accounting of affected entities.
Companies with exposed Siemens controllers can follow the advisory’s steps: inventory devices, remove public access, patch firmware, restrict network traffic and review remote-access protections. Organizations that rely on third-party cloud applications can review granted permissions and sign-in logs, and confirm that backup and response plans are usable. Users should start with email security, including a unique password and multifactor authentication where available. The near-term test will be whether organizations can reduce exposure and respond quickly as both attack tooling and vulnerability reporting evolve.
hardware security keys FIDO2 for business accounts
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did US agencies report about AI and industrial systems?
A joint advisory says actors used AI-generated scripts against internet-reachable Siemens S7 controllers. It warns of active activity but does not confirm that the reported targeting caused operational damage.
Did the Microsoft-reported campaign access customer data?
The supplied reporting does not establish that data was accessed. Microsoft says most targeting failed and that a successful sign-in does not necessarily indicate data access; ConnectWise said existing backup archives showed no indication of compromise.
Does AI make cyberattacks more successful?
The reports show AI being used to automate or speed parts of cyber activity, but they do not establish a general increase in success rates. Microsoft said most of the targeting it observed was unsuccessful.
What should organizations do now?
Prioritize asset inventories, patching, phishing-resistant sign-in, third-party app reviews, network separation for operational technology, and practised incident response. For industrial controllers, the advisory also recommends removing internet exposure and monitoring relevant traffic.
Is a coming “vulnerability tsunami” confirmed?
No. That phrase is an expert’s forecast, not a measured result. OpenAI also forecast a sharp rise in AI-found vulnerabilities, but the scale and timing remain uncertain.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
