Why AI-Driven Cyberattacks Demand A Faster Response

🔍 Read the full analysis: Why AI-Driven Cyberattacks Demand A Faster Response 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 separate cases of attackers using AI-assisted tools against industrial controllers and cloud services, putting a sharper focus on how quickly organizations can identify and contain intrusions. The accounts document active targeting and automated attack steps, but they do not show that AI makes every attempt successful: Microsoft said most of the targeting in its reported campaign failed.

A joint advisory issued August 18 by the NSA, CISA, FBI, Department of Energy and Environmental Protection Agency says actors used AI-generated scripts against internet-reachable Siemens S7 programmable logic controllers, or PLCs. These industrial computers help control equipment such as pumps, valves and production lines. The agencies said the activity involved reconnaissance and capability development targeting US critical infrastructure, and assessed that the actors were likely preparing for possible operational effects.

According to the advisory, actors found exposed controllers through internet scanning services and used scripts built on the open-source snap7 library, disguising them as monitoring tools. The agencies said this approach could reduce the expertise and time needed to attempt exploitation. They recommended basic safeguards, including removing PLCs from direct internet access, patching firmware, separating operational technology from corporate networks and monitoring for unauthorized control-system commands.

In an October 21 report, Microsoft Threat Intelligence linked the China-state-aligned actor Storm-0940 to use of ARTEX, an open-source agentic penetration-testing framework, in activity involving the ConnectWise Cloud Backup service. Microsoft said the actor tried to access more than 2,250 tenants using credentials associated with the application. It recorded more than 151,000 failed sign-in attempts from October 10 to 14, with a peak of 2,900 failures per minute. In several cases, Microsoft said, the time from initial targeting to privileged cloud access was under three hours.

Microsoft also cautioned that most of the targeting was unsuccessful and that use of an affected application or a successful sign-in did not, on its own, show that data had been accessed. ConnectWise said existing backup archives showed no indication of compromise. It suspended Cloud Backup, SaaS Security and Cloud Management while investigating.

Shorter Attack Timelines Test Defenses

The reports point to a practical change: automation can compress routine attack work, including scanning, repeated login attempts and adapting tools. That can leave defenders less time to spot suspicious activity and limit access. The evidence does not establish that attackers have gained a general ability to breach organizations at will; in Microsoft’s account, the majority of targeting attempts did not succeed.

Industrial systems and cloud services carry different risks. A compromised cloud account may expose organizational data or administrative access, while access to operational technology can affect physical processes. The government advisory’s concern is potential operational impact, not confirmation that such disruption occurred. For organizations, the response is to reduce exposure and prepare to act quickly, rather than treating AI as a substitute for familiar security controls.

AI can also help defenders find weaknesses. OpenAI said it found critical vulnerabilities in PHP that were patched on October 21, and forecast that AI-discovered critical flaws would “increase sharply in the coming months.” That is a forecast, not a count of flaws already found. If the pace of discovery rises, organizations may face greater pressure to prioritize and install patches while continuing to manage existing security work.

Two Reports, Different Attack Paths

The developments described in the source reports are not a single incident. The August 18 government advisory concerns attempts to identify and target exposed Siemens PLCs. Microsoft’s October report describes activity involving a trusted cloud application and a separate actor it identifies as Storm-0940. The details and potential consequences differ, so the cases should not be treated as evidence of one coordinated campaign.

Microsoft said Storm-0940 had been active since 2021 and assessed the actor as China-state-aligned. Its report says ARTEX can automate activity from reconnaissance through data exfiltration with limited human interaction. Separately, ConnectWise disclosed a compromise involving Cloud Backup on October 16 and suspended several services while investigating. The available source material does not establish that every targeted tenant suffered a breach.

A third example comes from Unit 42, which reported on July 30 that a Chinese-speaking actor connected the DeepSeek model to the Hermes Agent framework and directed it through Telegram. The supplied account says the agent searched GitHub after an initial exploit failed, illustrating how an automated workflow may adapt its next step. It does not provide enough detail here to establish the full outcome of that activity.

These reports support a narrower conclusion than claims of a new class of all-powerful attacks: AI is appearing in documented attack workflows, often alongside existing tools and methods. The change for defenders may be the volume and pace of attempts, rather than a wholly new kind of vulnerability.

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

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

Impact and Forecasts Remain Unsettled

The reports do not establish how many attempted intrusions led to confirmed access or what information, if any, was taken from affected cloud tenants. Microsoft says most targeting failed and warns that a successful sign-in does not necessarily mean data was accessed. ConnectWise said its existing backup archives showed no indication of compromise, while its investigation and service suspensions were continuing.

The government advisory describes activity against PLCs and assesses that actors were likely preparing for operational effects; it does not confirm that physical disruption occurred. The source material also cites an unnamed security leader predicting a “vulnerability tsunami” within months and estimating about 18 months to respond. That is an individual expert’s view, not a measured forecast or an agency finding. The scale, timing and practical effect of any increase in AI-assisted attacks remain unclear.

Investigations and Basic Controls

ConnectWise’s investigation and its review of the suspended services will help clarify whether the campaign led to customer data access. Microsoft’s report also underscores the need for organizations to review third-party app permissions, monitor unusual sign-in patterns and limit credentials that can reach multiple tenants.

For industrial operators, the agencies’ recommendations include inventorying controllers, applying firmware updates, blocking unnecessary internet access and separating operational technology from corporate networks. They also advise requiring multifactor authentication for remote access and monitoring controller traffic for commands outside approved change windows. Companies should test incident-response plans so teams know who can isolate affected systems and how to keep essential operations running.

Organizations do not need to wait for a new AI-specific product to take those steps. They can set patch deadlines, review external applications and practice response procedures. Individual users can begin with their email accounts by using phishing-resistant sign-in where available and reviewing account recovery options. The next evidence to watch is confirmed impact from ongoing investigations and whether future advisories document a measurable change in attack outcomes.

Key Questions

Have AI-assisted attacks caused confirmed damage to industrial systems?

The cited government advisory says actors targeted Siemens PLCs and were likely preparing for operational effects. The supplied material does not confirm physical disruption or damage.

Did the Storm-0940 campaign access data from every targeted tenant?

No. Microsoft said most targeting attempts were unsuccessful and that a successful sign-in does not necessarily mean data was accessed. ConnectWise said existing backup archives showed no indication of compromise while it investigated.

What does AI change about cyberattacks?

The reports describe AI-assisted scripting and agent frameworks used to automate steps such as reconnaissance and login attempts. They suggest that some tasks can happen faster; they do not show that every attack succeeds or that AI replaces other methods.

What should organizations do now?

Follow established controls: inventory and patch systems, remove industrial controllers from direct internet access, review third-party app permissions, require multifactor authentication for remote access and practice incident response. These steps address risks described in the reports without depending on predictions about future attack volume.

Source: ThorstenMeyerAI.com