It’s Worse Than We Thought: AI Has Taken The Clock Off Cyberattacks. What Companies And Users Should Do Now

🔍 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

Before you orderOffer from Amazon

Get the latest gadgets 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 documented cases of AI-assisted cyber activity, including scripts aimed at internet-reachable industrial controllers and an agentic framework used in attacks involving cloud tenants. The accounts show attackers applying AI to routine steps such as reconnaissance and exploitation, but Microsoft said most of the targeting it observed was unsuccessful, and some successful sign-ins did not establish that data had been accessed.

A joint advisory published August 18 by the NSA, CISA, FBI, Department of Energy and Environmental Protection Agency says threat actors used AI-generated exploitation scripts against Siemens S7 programmable logic controllers, or PLCs. These industrial computers can control equipment such as pumps and valves. The agencies said actors found exposed devices through internet scanning services and used scripts built with the open-source snap7 library, disguised as legitimate monitoring tools. They assessed the activity as likely preparation for possible operational effects, not proof that such effects occurred.

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 connected with a ConnectWise Cloud Backup compromise. Microsoft said the actor attempted sign-ins to more than 2,250 tenants using credentials associated with the cloud backup app. From October 10 to 14, Microsoft recorded more than 151,000 failed logons, peaking at 2,900 per minute; none of those attempts succeeded. In several attacks, Microsoft said, the time from initial targeting to privileged cloud access was under three hours.

Those figures require qualification. Microsoft said the majority of targeting was unsuccessful and that an affected app 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 while investigating. The supplied reporting also describes a July 30 Unit 42 account of an actor using a model with the Hermes Agent framework, but the available source text ends before it provides the details of what the agent did.

AI Compresses Routine Attack Work

The reported change is less about AI creating a wholly new kind of attacker and more about automating familiar attack steps. Scanning for exposed systems, trying credentials and adapting public tools can be time-consuming when done manually. Software that can run continuously may let an operator attempt those tasks at greater speed or across more targets, even when many attempts fail.

For organizations, that can narrow the time available to spot and contain an intrusion. Microsoft’s account of privileged access in under three hours illustrates why rapid detection and response matter; it does not establish that every AI-assisted attack moves at that speed. Industrial systems bring a separate risk: an internet-exposed controller could affect physical operations if an actor gains the ability to change its settings. The agencies described preparation for operational effects, rather than confirmed disruption.

There is also pressure on the defensive side. OpenAI said AI had identified critical PHP vulnerabilities patched on October 21 before attackers did, and forecast that AI-discovered critical flaws would increase sharply in the coming months. That is a company statement and a forecast, not a measured estimate of future vulnerabilities. If discovery does accelerate, organizations may face more frequent decisions about prioritizing and applying patches.

Three Reports, Different Evidence

The reports describe distinct forms of activity, not one unified campaign. The August 18 government advisory concerns scripts targeting Siemens S7 PLCs in sectors including manufacturing, energy, water, chemicals, food and agriculture, and commercial facilities. Its recommendations focus on limiting exposure: inventory controllers, update firmware, keep PLCs off the public internet, separate operational technology from IT networks and monitor industrial traffic.

Microsoft’s October 21 account concerns a third-party cloud application and tenant access. ConnectWise disclosed the Cloud Backup compromise on October 16, according to the source material. Microsoft attributed activity to Storm-0940 and described ARTEX as capable of coordinating activity from reconnaissance through data exfiltration with limited human interaction. These are Microsoft’s assessments; the reported scale of attempted logons should not be confused with the number of confirmed compromises.

The source article also relays a security leader’s expectation of a “vulnerability tsunami” within months and an estimate of roughly 18 months to hold on. Those statements are expert opinion, not a forecast supported by a stated measurement or a confirmed timeline. The evidence presently described supports concern about faster attack workflows, but not a precise timetable for how threats or defenses will change.

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

— Joint advisory by the NSA, CISA, FBI, DOE and EPA, published August 18

Access Does Not Prove Data Theft

The extent of any data access remains unclear from the supplied reporting. Microsoft explicitly said a successful sign-in or affected app does not necessarily mean data was accessed. ConnectWise reported no indication of compromise in existing backup archives, but the source material does not provide the final findings of its investigation or specify the full scope of any exposure.

The agency advisory describes activity against PLCs and assesses that actors may be preparing for operational effects; it does not report confirmed physical disruption. The source also offers no baseline for measuring how much AI has increased attack volume or success rates. Claims that a vulnerability surge is imminent, or that organizations have a specific amount of time to prepare, remain forecasts rather than established facts.

Investigations and Basic Defenses

ConnectWise’s investigation and any further findings from Microsoft or the company will clarify whether tenant access led to data exposure. Organizations using affected services should follow vendor updates and review their own access and audit records. The source material does not give a timeline for a final investigation report.

For industrial operators, the agencies recommend checking controller inventories and firmware, removing public internet access to PLCs, separating IT and operational technology networks, requiring multi-factor authentication for remote access and monitoring for unexpected write activity. The recommendations also include blocking TCP port 102 at the perimeter and using one-way gateways for historian connections where appropriate.

For companies more broadly, the immediate response is to review third-party app permissions, set patch deadlines, use phishing-resistant sign-in where available, and practise incident response. The reporting’s central warning is about less time between targeting and access, not proof that every attack is more effective. Users can start by securing their email account, a common route to password resets and other services.

Key Questions

What did US agencies say attackers did with AI?

The agencies said actors used AI-generated scripts against internet-reachable Siemens S7 programmable logic controllers. They described the activity as active, and assessed that it could be preparation for operational effects; the advisory does not confirm physical disruption.

Did Microsoft confirm that 2,250 organizations were compromised?

No. Microsoft said the actor attempted sign-ins to more than 2,250 tenants. It also said most targeting was unsuccessful and cautioned that an affected app or successful sign-in does not necessarily mean data was accessed.

What does the report say about the ConnectWise investigation?

ConnectWise said existing backup archives showed no indication of compromise and that it suspended Cloud Backup, SaaS Security and Cloud Management while investigating. The supplied material does not give the investigation’s final results.

Is a surge in AI-discovered vulnerabilities confirmed?

No. OpenAI said it expected AI-found critical vulnerabilities to increase sharply in the coming months. That is a forecast, not a confirmed count or a measured estimate of how many flaws will be found.

What should organizations do now?

Review exposed systems and third-party app access, apply security updates on defined timelines, require strong authentication for remote access, separate operational technology from business networks, and practise incident response. The government advisory also recommends keeping PLCs off the public internet and monitoring industrial traffic.

Source: ThorstenMeyerAI.com