Why Faster AI-Driven Cyberattacks Call For Better Preparation

🔍 Read the full analysis: Why Faster AI-Driven Cyberattacks Call For Better Preparation on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get hardware and tech essentials delivered free with Prime

• Fast, free delivery on millions of items
• Prime Video, Amazon Music and more included
• Member-only deals all year

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

US agencies and Microsoft have reported AI-assisted cyber operations targeting industrial controllers and cloud tenants, including the use of generated scripts and an automated penetration-testing framework. The reports do not show that every attempt succeeded, but they describe how automation can speed reconnaissance and repeated login attempts, raising the stakes for organizations that rely on industrial or cloud systems.

A joint advisory published August 18 by the NSA, CISA, FBI, Department of Energy and EPA says threat actors used AI-generated exploitation scripts against internet-reachable Siemens S7 programmable logic controllers, or PLCs. PLCs control equipment in facilities such as manufacturing plants and water systems. The agencies said the activity involved reconnaissance and capability development and assessed that it was likely preparation for operational effects against critical infrastructure.

Microsoft Threat Intelligence reported October 21 that Storm-0940, which Microsoft describes as aligned with the Chinese state, used the open-source ARTEX agentic framework after a compromise involving ConnectWise Cloud Backup. Microsoft said the actor used credentials associated with the cloud backup application to attempt sign-ins to more than 2,250 customer tenants. In a password-spraying campaign from October 10 to 14, Microsoft recorded more than 151,000 failed logons, peaking at 2,900 per minute; Microsoft said none of those attempts succeeded.

Microsoft said some attacks reached privileged cloud access in under three hours, while stressing that most targeting was unsuccessful. It also cautioned 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 that it suspended Cloud Backup, SaaS Security and Cloud Management during its investigation.

Automation Compresses Attack Timelines

The reports point to a practical change in the pace of cyber operations, rather than evidence of an all-powerful or autonomous attacker. AI tools can help carry out familiar steps—scanning for exposed systems, adapting scripts or trying credentials—at a scale and speed that may leave defenders less time to identify unusual activity and respond. Microsoft noted that agents can work continuously; that is a description of capability, not proof that every operation is effective.

For operators of critical infrastructure, an exposed or poorly segmented controller can have consequences beyond data theft, since PLCs affect physical processes. In cloud environments, a trusted third-party application can create a route into customer tenants if its credentials are compromised. These cases make inventory, access controls and monitoring important across both internal systems and supplier relationships.

AI also changes the workload on defenders. OpenAI said AI identified critical PHP vulnerabilities that were patched on October 21 before attackers exploited them, and forecast that AI-found critical vulnerabilities would rise sharply in the coming months. That is a company forecast, not a measured count of future flaws. If vulnerability discovery accelerates, organizations may face more pressure to assess and apply fixes quickly, while still verifying that patches do not disrupt essential services.

Three Reports, Different Evidence

The August advisory describes activity against Siemens S7 PLCs, including models from the S7-200 through S7-1500. The agencies said attackers used internet scanning services to find exposed devices and scripts built with the open-source snap7 library, disguised as monitoring tools. Their advisory called the activity an active threat and said AI use could reduce the technical expertise and time needed for some tasks. The agencies recommended keeping PLCs off the public internet, patching firmware, separating operational technology from business IT and monitoring S7 traffic.

Microsoft’s October report concerns a separate incident involving ConnectWise Cloud Backup. ConnectWise first disclosed the compromise on October 16. Microsoft attributed the subsequent tenant targeting to Storm-0940 and described ARTEX as a multi-agent framework capable of tasks from reconnaissance through data exfiltration with limited human interaction. The report’s figures distinguish failed attempts from successful access: the 151,000-plus logons Microsoft recorded failed, while separate attacks reportedly reached privileged access in some cases.

A July 30 report from Palo Alto Networks’ Unit 42 described another case in which a Chinese-speaking actor connected the DeepSeek model to the Hermes Agent framework and directed it through Telegram. Unit 42 said the agent responded to a failed exploit by searching GitHub for alternatives. The cases do not establish that AI alone planned or carried out attacks; they document people using AI tools within broader operations.

“This is not a theoretical risk—it is an active threat.”

— Joint Cybersecurity Advisory by the NSA, CISA, FBI, Department of Energy and EPA

Impact and Forecasts Still Unsettled

The full consequences of the ConnectWise-related activity remain unclear. Microsoft said some attacks achieved privileged cloud access, but also warned that successful sign-in does not necessarily mean data was accessed. ConnectWise said its existing backup archives showed no indication of compromise; the source material does not provide a final account of the investigation or establish whether customer data was taken.

The government advisory says agencies assess the PLC activity as likely preparation for operational effects, but it does not confirm that the reported activity caused physical disruption. The source material also does not quantify how much AI contributed to the attacks compared with other tools or human decisions. Predictions of a near-term “vulnerability tsunami” are expert opinion, not an established measurement or timeline.

Security Reviews and Investigations Continue

Organizations affected by the ConnectWise incident will need to follow updates from ConnectWise and Microsoft as investigations clarify what access occurred and whether any information was reached. Operators of industrial systems can compare their environments with the joint advisory’s recommendations, including checking for internet-exposed PLCs, reviewing remote access and monitoring controller traffic for unexpected writes.

For companies more broadly, the recommendations in the source reports are conventional but operationally demanding: maintain an accurate asset inventory, set patch deadlines, review third-party applications and permissions, use phishing-resistant sign-in where possible, and separate operational technology from corporate networks. Incident-response plans should be exercised, not merely documented. Individuals can start by securing their email accounts with multifactor authentication and checking recovery options, since email often controls access to other services.

Further reporting should establish the final findings from the ConnectWise investigation and whether agencies or threat researchers publish additional evidence about the activity. Until then, the confirmed record supports preparation for faster attempts—not assumptions that every attack succeeds or that AI makes breaches inevitable.

Key Questions

What did US agencies report about AI and industrial systems?

A joint advisory published August 18 said actors used AI-generated scripts against internet-reachable Siemens S7 controllers. The agencies described reconnaissance and capability development and assessed that the activity was likely preparation for possible operational effects.

Did Microsoft say the attacks accessed customer data?

No. Microsoft reported tenant targeting and said some attacks reached privileged cloud access, but cautioned that an affected application or successful sign-in does not necessarily mean data was accessed. ConnectWise said its existing backup archives showed no indication of compromise while it investigated.

What should organizations do in response?

Based on the cited advisories, organizations should inventory systems, apply firmware and software patches, restrict internet exposure, strengthen remote-access authentication, review third-party app permissions and practice incident response. Industrial operators should also separate operational technology from business networks and monitor controller traffic.

Does this prove AI can carry out cyberattacks without people?

No. The reports describe attackers using AI tools and agentic frameworks within operations. They do not establish that AI independently planned or executed the activity, or that AI use made every attempt successful.

Source: ThorstenMeyerAI.com