What AI Means For The Speed Of Cyberattacks—and Your Defenses

🔍 Read the full analysis: What AI Means For The Speed Of Cyberattacks—and Your Defenses on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

US agencies and Microsoft have reported separate cases of AI-assisted cyber activity, including scripts targeting industrial controllers and an AI-agent framework used in attacks involving cloud services. The disclosures, alongside OpenAI’s warning that AI-found vulnerabilities are expected to increase, point to faster attack and patch cycles—not proof of an autonomous, all-powerful attacker.

A joint advisory issued August 18 by the NSA, CISA, FBI, Department of Energy and Environmental Protection Agency says actors used AI-generated exploitation scripts against internet-reachable Siemens S7 programmable logic controllers. These industrial computers help operate facilities such as manufacturing plants, energy and water sites. The agencies say the scripts were disguised as monitoring tools and built using the open-source snap7 library. They assess the activity as likely preparation to cause operational effects, but the advisory does not say that such effects occurred.

In an October 21 report, Microsoft Threat Intelligence said the China-state-aligned actor it calls Storm-0940 used ARTEX, an open-source agentic penetration-testing framework, in activity linked to a ConnectWise Cloud Backup compromise. Microsoft said the actor attempted sign-ins to more than 2,250 tenants after obtaining credentials for the cloud backup app. It recorded more than 151,000 failed logons during a password-spraying campaign from October 10 to 14, with a peak of 2,900 failures per minute. None of those logons succeeded, Microsoft said.

Microsoft also reported that in several attacks, the time from initial targeting to privileged cloud access was under three hours. It cautioned that most targeting was unsuccessful and that an affected app or successful sign-in does not by itself show that data was accessed. ConnectWise, which disclosed the compromise October 16, said existing backup archives showed no indication of compromise and that it suspended Cloud Backup, SaaS Security and Cloud Management while investigating.

Faster Attacks, Tighter Response Windows

The reports matter because automation can compress routine attack steps—reconnaissance, password spraying and adapting exploit code—into shorter periods. That can leave defenders less time to detect activity and respond, especially when attackers use credentials or trusted software relationships that already provide access. The Microsoft figures are evidence of activity at scale, not evidence that most attempts worked.

AI also affects the defensive side. OpenAI said it had used AI to identify critical vulnerabilities in PHP that were patched October 21, before attackers exploited them, and forecast that AI-found critical flaws would increase sharply in the coming months. That is a forecast, not a measured count of future flaws. If discoveries rise, organizations may face more urgent patch decisions, including for systems that cannot be updated without operational planning.

The practical implication is not that organizations need to buy an AI product. The cited advisories emphasize fundamentals: know what systems and applications are exposed, limit access, review third-party connections, apply security updates and rehearse incident response. For industrial operators, keeping controllers off the public internet and separating operational technology from business networks can reduce exposure.

Three Disclosures, Different Evidence

The developments describe distinct activity and should not be treated as one coordinated campaign. The August advisory concerns Siemens S7 controllers and AI-generated scripts. Microsoft’s October report concerns Storm-0940, ARTEX and activity associated with a cloud backup application. OpenAI’s statement concerns AI-assisted vulnerability discovery and future expectations. Each source supports a different part of the broader picture.

Microsoft describes Storm-0940 as active since 2021 and ARTEX as a multi-agent framework capable of supporting stages from reconnaissance to data exfiltration with limited human interaction. That capability does not establish that an agent independently chose every target or completed every step. In the reported activity, most targeting attempts did not succeed, and Microsoft explicitly cautioned against equating access attempts with data theft.

The source material also cites a Unit 42 report describing an actor using the DeepSeek model with the Hermes Agent framework and Telegram. The account says the agent searched GitHub after an initial exploit failed, but the available material does not provide enough detail here to establish the operation’s outcome or wider impact. Claims of a coming “vulnerability tsunami” and an 18-month window are an unnamed security leader’s opinion, not an independently measured forecast.

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

— The joint NSA, CISA, FBI, Department of Energy and EPA advisory

Impact and Scale Still Unclear

It is not clear whether the reported controller activity caused operational disruption, or how many facilities were exposed or affected. The advisory describes likely preparation for operational effects, not confirmed damage. Microsoft says its observed targeting was mostly unsuccessful and that a successful sign-in alone does not establish data access; ConnectWise said its backup archives showed no indication of compromise.

The reports also do not establish how widely attackers are using AI-generated code or agent frameworks across incidents, or how much AI changes success rates compared with conventional automation. OpenAI’s forecast gives no public numerical estimate in the material cited here. The unnamed expert’s prediction of a near-term “vulnerability tsunami” and roughly 18 months to respond remains an opinion, not a verified timeline.

Patch and Access Reviews Continue

ConnectWise said it suspended the affected services while investigating; the cited material does not provide a final investigation result. Organizations using those services should follow vendor updates and review their own access and incident records rather than infer compromise from an attempted sign-in alone.

For industrial operators, the federal advisory recommends inventorying controllers, patching firmware, removing internet exposure, blocking TCP port 102 at the perimeter, separating IT and operational technology networks, requiring multifactor authentication for remote OT access and monitoring S7 write activity outside approved maintenance windows. For other organizations, the immediate work is to review third-party app permissions, tighten sign-in protections, prioritize patches and practise response procedures. Users can begin by securing their email accounts with multifactor authentication and a unique password.

Further clarity will depend on investigation updates, evidence of any confirmed impact and whether future security reports document more AI-assisted activity. For now, the available evidence supports a measured conclusion: AI can speed parts of cyber operations, while basic controls remain relevant to both the attacks and the defense.

Key Questions

Did AI cause a confirmed cyberattack on industrial facilities?

US agencies said actors used AI-generated scripts against Siemens S7 controllers and assessed the activity as likely preparation for operational effects. The cited advisory does not confirm that those effects occurred.

Did the Storm-0940 activity compromise thousands of cloud tenants?

Microsoft said the actor attempted sign-ins to more than 2,250 tenants after obtaining cloud backup app credentials. It also reported that the majority of targeting was unsuccessful. An attempted sign-in does not establish that a tenant’s data was accessed.

What does AI change about cyberattacks?

The reports describe AI and agent frameworks helping with familiar tasks such as reconnaissance, password spraying and exploit-script development. They indicate potential speed and scale advantages, but do not show that attacks are automatically successful or fully independent of human direction.

What should organizations do now?

Organizations should inventory exposed systems, patch firmware and software, limit internet access to industrial controllers, review third-party app permissions, require multifactor authentication for remote access and practise incident response. These steps align with recommendations in the cited advisory and address the access paths described in the reports.

Is a surge in AI-found vulnerabilities confirmed?

No. OpenAI said it expects AI-found critical vulnerabilities to increase sharply in coming months, but that is a forecast. The cited material does not give a numerical estimate or establish how many such discoveries will occur.

Source: ThorstenMeyerAI.com