🔍 Read the full analysis: What Users Should Do As AI Changes The Speed Of Cyberattacks on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
US agencies and Microsoft have described separate incidents in which attackers used AI-generated scripts or an agentic framework, though Microsoft says most attempts in its reported campaign failed. AI can shorten the time needed for routine attack steps, while defenders are also using AI to find vulnerabilities. Users should secure email and other high-value accounts; companies should prioritize patching, access controls, third-party reviews and incident response.
US agencies and Microsoft have reported separate cases of AI-assisted cyber activity, including scripts aimed at industrial controllers and an agentic framework used in attacks on cloud tenants. The reports show how automation can speed up familiar attack methods, but they do not establish that AI made the attacks broadly successful: Microsoft said most of the targeting in its case failed.
A joint advisory published August 18 by the NSA, CISA, FBI, Department of Energy and EPA said actors were using AI-generated exploitation scripts against internet-reachable Siemens S7 programmable logic controllers, or PLCs. PLCs are industrial computers used to control equipment and processes. The advisory described the scripts as disguised monitoring tools built with the open-source snap7 library, and said the activity involved reconnaissance and capability development against US installations.
On October 21, Microsoft Threat Intelligence reported that Storm-0940, which Microsoft describes as a China-state-aligned actor, used the open-source ARTEX multi-agent framework in activity linked to a ConnectWise Cloud Backup compromise. Microsoft said the actor used credentials for the cloud backup application to attempt sign-ins to more than 2,250 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 that in several attacks, the time from initial targeting to privileged cloud access was under three hours. It also cautioned that most targeting attempts were unsuccessful and that an affected application or successful sign-in does not necessarily mean 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.
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
Automation Compresses Attack Timelines
The reports point to a practical change: automation can reduce the time and effort needed for scanning, credential attempts and adapting scripts. That does not make every operation effective, and the Microsoft figures include a large set of failed attempts. But a shorter interval between initial targeting and privileged access can leave defenders less time to identify unusual activity and respond.
The risks extend beyond cloud accounts. The government advisory concerns industrial control systems, where unauthorized access could affect physical operations. The agencies assessed the activity as likely preparation to cause operational effects, not as proof that such effects had occurred. Their recommendations focus on limiting exposure and separating operational technology from ordinary business networks.
AI also affects the defensive workload. OpenAI said AI had identified critical PHP vulnerabilities that were patched on October 21 and forecast that similar AI-found vulnerabilities would increase sharply in coming months. That is a company forecast, not a measured estimate of future incidents. If vulnerability discovery accelerates, organizations may face more pressure to verify and apply fixes promptly.
hardware security key FIDO2 USB-C phishing resistant MFA
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Two Documented Attack Paths
The industrial-control warning and the cloud campaign are distinct cases, not evidence of a single coordinated operation. In the first, agencies described AI-generated scripts targeting Siemens PLCs found through internet scanning. They identified manufacturing, energy, water, chemical, food and agriculture, and commercial facilities among the sectors of concern.
In the second, Microsoft described a trusted software relationship being used as an entry point: credentials associated with a cloud backup application were used to attempt access to customer tenants. Microsoft said ARTEX can operate across stages from reconnaissance to data exfiltration with limited human interaction. The account is Microsoft’s assessment; ConnectWise separately said its investigation had not found indications of compromise in existing backup archives.
A separate July 30 report from Palo Alto Networks’ Unit 42 described an actor using the DeepSeek model with the Hermes Agent framework and steering it through Telegram. Unit 42 said an agent performed hundreds of hours of manual targeting analysis in minutes. That example and the other reports illustrate automation of ordinary attack tasks, rather than establishing that AI has created a wholly new class of attack.
“This is not a theoretical risk—it is an active threat.”
— Joint advisory by the NSA, CISA, FBI, Department of Energy and EPA, published August 18
password manager family business passkeys MFA
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
What the Reports Do Not Establish
The reports do not show that every AI-assisted attempt succeeded or that AI alone caused the outcomes described. In Microsoft’s account, most targeting failed, and the company said a successful sign-in or affected application does not by itself establish that customer data was accessed. ConnectWise’s statement about existing archives is its own finding; the full scope and results of the investigation were not provided in the source material.
The agencies described likely preparation to cause operational effects against critical infrastructure, but the advisory does not confirm that the activity caused physical disruption. Nor do the reports quantify how frequently AI is used across all cyberattacks or show how much faster attacks are on average. Predictions of a coming wave of vulnerabilities or worsening conditions remain forecasts, not established outcomes.
small business firewall VPN intrusion prevention system
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Prioritize Account and Network Defenses
For individuals, the immediate starting point is the email account, which can be used to reset passwords elsewhere. Use a unique password stored in a password manager and turn on phishing-resistant multi-factor authentication where available. Review recovery details and active sessions, and be cautious about unexpected sign-in prompts or requests to approve access to an application.
Organizations should focus on basic controls identified in the government advisory and the reports: keep an inventory of internet-facing systems and industrial controllers, set patch deadlines, review third-party applications and their permissions, and require strong authentication for remote access. For operational technology, agencies recommend removing PLCs from direct internet access, separating IT and OT networks, and monitoring for unexpected controller writes. They also advise blocking TCP port 102 at the perimeter where applicable.
Security teams should rehearse incident response and decide in advance how they will revoke credentials or app access, isolate affected systems and notify customers. The next confirmed milestones are the outcome of ConnectWise’s investigation and any further advisories or vulnerability disclosures. Until those arrive, reported incidents should be distinguished from forecasts about how the threat may develop.
network security monitoring appliance industrial OT PLC
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Does the reporting show that AI cyberattacks are succeeding more often?
No. Microsoft said most of the targeting it described was unsuccessful, and it recorded more than 151,000 failed logons in one campaign, with none succeeding. The reports document AI use and faster activity, not a general increase in success rates.
What should an individual do first?
Secure your email account with a unique password and phishing-resistant multi-factor authentication if available. Review account recovery information and connected applications, since email can be used to reset access to other services.
What did US agencies say about industrial controllers?
A joint advisory said actors used AI-generated scripts against internet-reachable Siemens S7 PLCs. The agencies recommended inventorying controllers, patching firmware, removing direct internet exposure, separating IT and OT networks, and monitoring controller traffic.
The source material does not confirm that. Microsoft said an affected application or successful sign-in does not necessarily mean data was accessed. ConnectWise said existing backup archives showed no indication of compromise while it investigated.
Are forecasts of a coming vulnerability surge confirmed?
No. OpenAI forecast that AI-found critical vulnerabilities would increase sharply in coming months, while a security leader’s prediction of a near-term “vulnerability tsunami” was an expert opinion. Neither is a measured count of future vulnerabilities.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
