Issue 53 · Nov 16–22, 2026Agent Age News magazine · Other editions · Search · Method · HU / ENG

Incidents and influence attempts in the age of agents

Narrative · ▲ at peak

The AI agent now attacks independently

All narratives · 24 supporting stories · 0 from Hungary and the region

„Attackers use AI agents to fully automate cyberattacks from reconnaissance to data theft, enabling smaller actors to run state-level operations.”

Counter-narrative

The label 'first fully autonomous attack' keeps coming up, but according to several reports, AI agents are not yet capable of fully autonomous attacks: they are accelerator tools, and human involvement is also required. This is represented by some reports and critics of the Sysdig case.

Voices

Anthropic, Google, Microsoft, FBI, Sysdig, professional security press

Analysis

According to the narrative, agents have appeared in every stage of the attack chain, and the barrier to entry has collapsed. It is fed by AI labs' threat reports and security firms' case write-ups, which award the label "first" several times in a row. Labs have an interest in acting as responsible gatekeepers, and security vendors in selling AI-based defense. The weak point is that there is no commonly accepted definition of the boundary between "autonomous" and "AI-assisted." The debate over the Sysdig case and the contradictory February reports show how fluid the basic concept is.

Evidence: moderate — Rests on concrete cases (Mexico 150 GB, Taiwan 85 accounts, PaperCut 440 servers, 395 organisations according to another report). These are mostly disclosed by the companies that sell the model or the security product, and the degree of autonomy is rarely verified independently. The FBI and Google warnings are partly forecasts about capability, not incidents that have happened.

What's at stake: The stake is whether defense aligns with real automation or a marketing concept.

Narrative sub-threads

The narrative splits into 5 branches; a story may belong to several branches, and some may not appear in the main narrative's list.

→ lasting · evidence: strong

State-sponsored espionage campaigns with autonomous agents

Actors linked to China, Russia, and other countries are running espionage operations against government and corporate targets using AI agents, from reconnaissance to data collection. According to Anthropic, China-linked attackers used agents against approximately thirty large corporations and government agencies. In Taiwan, autonomous agents breached 85 government accounts. CrowdStrike states that Chinese state hackers targeted more than 200 organizations with ARTEX.

It concerns state-sponsored or state-linked espionage groups and government victims, not financially motivated crime. The evidence consists of vendor threat reports and government statements.

9 stories · 0 of them Hungarian and regional
▲ at peak · evidence: strong

Small players at state scale

Individual hackers and small criminal groups are breaching hundreds of organizations using commercial AI tools, sometimes within hours. According to Anthropic, this previously required state-level capabilities. Examples: 600 firewalls, the PaperCut servers of 395 organizations, 2,250 Microsoft 365 tenants breached in hours, and 150 GB of Mexican government data.

The actor here is an individual or small group. The mechanism involves the abusive use of generative AI such as Claude, Cursor, DeepSeek, or others. The evidence points to specific mass intrusions.

12 stories · 0 of them Hungarian and regional
→ lasting · evidence: moderate

AI-written malware and automated ransomware

Attackers use AI to generate malware and exploits, such as VoidLink, the Hive Slopoly malware, and a zero-day exploit detected by Google. According to Sysdig, an agent fully automated a ransomware attack; this claim is disputed. One report states that agents can reduce ransomware intrusion time to under ten hours.

The focus is on automating attacker tools and ransomware chains, not the identity of the actor. Evidence includes malware analysis and incident descriptions.

8 stories · 0 of them Hungarian and regional
↗ rising · evidence: moderate

Authority alerts: ARTEX and forecasts

The FBI and Google warn that attackers using the open-source ARTEX testing tool and other AI platforms can gain privileged access within three hours in some breaches. Google says fully autonomous attacks could appear within three months. The UK NCSC considers AI-based ransomware attacks on critical infrastructure extremely likely by 2028.

It relies on law enforcement, regulatory, and manufacturer warnings and forecasts, not on a single incident. The ARTEX tool and agents building exploits from patches are at the center.

7 stories · 0 of them Hungarian and regional
↘ fading · evidence: medium

Contested autonomy and excessive fears

Sources say agents are not yet capable of fully autonomous attacks. The 'first' AI-driven cases are disputed regarding both human involvement and attribution. According to VulnCheck, only 1% of vulnerabilities discovered by AI were exploited in live attacks.

It carries counter-evidence and a skeptical assessment, questioning the evidence of other narrative branches.

5 stories · 0 of them Hungarian and regional

Sub-thread stories

State-sponsored espionage campaigns with autonomous agents

Voices: Anthropic, Google, Taiwanese government, CrowdStrike

Nov 17, 2025Anthropic says China-linked actors used AI agents in cyber espionage campaignsecurity leaderDefences must match the pace of an AI-accelerated, partly autonomous attack chain.security operationsAutomating detection and response is increasingly important against machine-speed attacks.public-sector decision-makerGovernment bodies are also among the targets mentioned in the report.journalist and researcherThe case is being analysed as the first AI-driven campaign presented as large-scale.Nov 24, 2025Anthropic discloses AI-orchestrated cyber-espionage campaign; Congress calls CEO to testifysecurity leaderThey must allow for the possibility that individual steps of an attack can be carried out by an AI agent, and adapt their defence model and detection accordingly.public-sector decision-makerThe congressional hearing signals that AI-driven espionage has become a question of regulation and liability.journalist and researcherAnthropic's account is the first documented case of AI-orchestrated attacks; its details still need to be verified from an independent source.Feb 12, 2026Google: state-backed groups used Gemini AI for reconnaissance and attack supportsecurity leaderAttackers are building generative AI into reconnaissance and attack planning, and the threat model needs to account for this.journalist and researcherGoogle's statement is a documented, vendor-confirmed example of AI misuse.Jul 27, 2026Hackers used an autonomous AI agent to spy on Thailand's finance ministrysecurity leaderAgent-based attacks prepare defenders for fast, partly automated attack chains.journalist and researcherThe case is a documented example of attackers using AI agents.Aug 11, 2026China-linked hackers attack Taiwan's government systems with autonomous AI agentssecurity leaderBecause of autonomous, open-source agent chains, the threat model must account for the speed of reconnaissance and account compromise.security operationsAnomaly monitoring of government accounts and protection of credentials come to the fore.public-sector decision-makerThe campaign shows that autonomous AI tools can now be used against state systems.journalist and researcherThe technical details of the open-source autonomous attack chain need analysis and verification.Aug 17, 2026Suspected Chinese hackers use autonomous AI tools to steal 2,500 personnel records from Taiwanese government systemsecurity leaderIn this case the attackers used autonomous AI tools, so defences must account for faster, automated operations.public-sector decision-makerGovernment personnel data emerged as a target, which bears on the protection of state data.journalist and researcherThe case may be among the first documented examples of AI-driven attacks; the claim about the suspected actor should be treated as sourced.Sep 11, 2026Russian state-sponsored hackers reportedly used Claude to rebuild detected malwaresecurity leaderState actors' use of AI is also a factor in defence planning for the region near Ukraine.security operationsRapidly rebuilding detected malware with AI can shorten the period in which signature-based defences remain effective.journalist and researcherThe attribution rests on Anthropic's claim; the excerpts contain no independent confirmation.Oct 30, 2026Chinese state-backed hackers used AI-powered ARTEX to target over 200 organizations, CrowdStrike sayssecurity leaderAttack paths parallelized by agents increase the speed of attacks.security operationsBLOODALCHEMY indicators should be added to detection.Nov 11, 2026Kazakhstan-linked hackers use autonomous AI agent Hermes to steal 2,000 Tajikistan government documentssecurity leaderEspionage carried out with autonomous agents can speed up data theft.public-sector decision-makerProtecting government document repositories against AI-based attacks is also an open question.

Small players at state scale

Voices: Anthropic, Amazon, GreyNoise, CrowdStrike

Feb 20, 2026AWS says AI-augmented threat actor accesses FortiGate devices at scalesecurity leaderAI increases attackers' ability to scale, so rapid patching and reduced exposure are key to perimeter defence.security operationsFortiGate configuration, exposure and authentication should be reviewed urgently.Feb 23, 2026Amazon says Russian-speaking hackers used generative AI to compromise 600 firewallssecurity leaderAI can also increase the scale of attacks by less skilled attackers.security operationsAccess controls and configuration of perimeter firewalls need to be rechecked.Feb 26, 2026Hackers reportedly used Anthropic's Claude to steal 150 GB of Mexican government datasecurity leaderThe case suggests that commercial AI tools can also speed up attackers' work.public-sector decision-makerProtecting government data now has to account for AI-assisted intrusion.Mar 4, 2026Hackers used Claude Code to build attack tools and steal data from Mexican government agenciessecurity leaderBecause coding agents are being used to develop attacks, abuse detection and pattern monitoring for agents are important.public-sector decision-makerAn AI-assisted attack on government data sets also raises data protection risks.Aug 3, 2026China-based hacker uses DeepSeek agent against a security firm's networksecurity leaderThe pace and method of attacks carried out by autonomous agents pose a new detection challenge.security operationsLook for signs of proxyjacking, such as unusual outbound traffic and relay services.journalist and researcherThe attackers' use of an AI agent provides a documented example for the topic.Aug 27, 2026Russian-speaking criminals used Cursor coding assistant to breach a Belgian chemicals firm and at least six otherssecurity leaderAI coding assistants have become part of attackers' toolkits, which affects defence planning.developerThe risk of attackers abusing widely used AI developer tools must be taken into account.business leaderChemicals and other industrial companies can also be targets of AI-assisted intrusions.Sep 8, 2026Autonomous AI agents compromise thousands of credentials in under six hourssecurity leaderAttacks that scale without human intervention require rethinking credential management and detection speed.security operationsBecause credential theft is fast and automated, rotating secrets and monitoring for anomalies are urgent.Sep 10, 2026Anthropic: AI lets smaller actors run state-level hacking campaignssecurity leaderThe report says AI appears in attacks at varying levels of autonomy, which affects threat modelling.security operationsAnthropic says attacker sophistication no longer depends on the size of the actor, so detection assumptions need rethinking.journalist and researcherThe report gives details of cases ranging from cyberattacks to influence operations, surveillance, biology and weapons building.Sep 10, 2026Hundreds of AI agents compromise PaperCut servers, GreyNoise sayssecurity leaderAttacks automated with AI agents speed up exploitation relative to the patching window.security operationsPatching the PaperCut vulnerabilities is urgent, as the campaign uses agents to scan and attack at scale.Sep 22, 2026Chinese-speaking hacker used AI agents to breach 100 companies and steal card datasecurity leaderMass, agent-automated attacks make protecting card-data systems urgent.business leaderVictims can include smaller firms, where defenses are weaker.Oct 8, 2026CrowdStrike suspects 26-year-old in China used Claude Code against South Korean bankssecurity leaderThe attribution rests on a single vendor's claim and is disputed by those involved, so it should be treated with caution.developerGiven the misuse of coding agents, provider logging and abuse detection are key issues.journalist and researcherUsing AI provider logs as attribution evidence raises new methodological questions.Oct 22, 2026China-linked AI agent breached 2,250 Microsoft 365 tenants within hours, bypassing MFAsecurity leaderMFA alone did not protect in this scenario, so tenant configurations should be reviewed.security operationsMicrosoft 365 tenant logs and access rights should be reviewed.

AI-written malware and automated ransomware

Voices: IBM, Google, Sysdig, DIVD

Jan 19, 2026VoidLink cloud malware shows signs of being AI-generatedsecurity leaderAI is speeding up malware development, so detection assumptions for cloud environments should be reviewed.security operationsBehaviour-based detection becomes more important against fast-changing malware that targets the cloud.journalist and researcherThe markers that identify AI-generated code are a research area in their own right.Mar 16, 2026IBM links suspected AI-generated "Slopoly" malware to Hive0163 ransomware attackssecurity leaderAttackers can build malware faster and more cheaply with AI, which may challenge current detection assumptions.security operationsDetection should also rely on behavioural signals, not only on known signatures.journalist and researcherIBM's claim needs independent verification; samples and related details can be examined by other researchers.May 11, 2026Google detects AI-developed zero-day exploit tied to cybercriminal groupsecurity leaderAI-accelerated vulnerability research and exploit development mean shorter defence windows, which should be factored into threat models and patch priorities.security operationsBecause of the reported bypass of two-factor protection, it is worth reviewing how authentication processes are monitored.developerBecause AI can uncover flaws, code auditing and rapid patching matter even more.journalist and researcherThe case is a defender-documented example of AI-assisted exploit development that researchers should follow.Jul 2, 2026Researchers say AI agent ran ransomware attack on database via Langflow flaw without human involvementsecurity leaderAutomating the attack chain can shorten the time between discovery and extortion.security operationsLangflow instances need to be patched and isolated, and databases need strict access controls and backups.developerExposed instances of AI workflow frameworks are entry points for attackers.journalist and researcherThe researchers say the attack ran without human direction; no independent source has yet confirmed this.Sep 3, 2026AI agents help cut ransomware intrusions to under 10 hourssecurity leaderThe window for detection and response narrows significantly.security operationsAutomated response and alert prioritization are needed against fast intrusions.Sep 29, 2026DIVD says Zammad zero-days enabled AI-driven network breachsecurity leaderAn attack carried out by an automated agent allows faster exploitation, and exposure of support systems is a particular concern.security operationsZammad instances need to be patched and checked; vulnerability-disclosure organizations can be targets too.journalist and researcherAn attack on organizations that uncover vulnerabilities signals a risk to the research ecosystem.Oct 5, 2026Autonomous AI agent chains two Zammad zero-days to breach Dutch vulnerability disclosure institute DIVDsecurity leaderDefence plans should also account for the speed of automated, agent-based attacks.security operationsUpdating Zammad instances and reviewing the exposure of helpdesk systems may be urgent.developerAgents can chain vulnerabilities together at machine speed.Oct 31, 2026New AI-driven Windows malware sends stolen data to Claude to rank victims' assets by valuesecurity leaderAttackers are using AI to automate the assessment of victims as targets.security operationsUnusual outbound traffic to AI services can be a detection signal.journalist and researcherRanking victims by potential loss is a new technique, and analysis of it may be ongoing.

Authority alerts: ARTEX and forecasts

Voices: FBI, Google, NCSC, Microsoft

Mar 7, 2026Microsoft says attackers use AI at every stage of cyberattackssecurity leaderThreat models should account for AI-assisted attacker activity at every stage of the chain.journalist and researcherMicrosoft's claim gives a starting point for documenting and independently verifying attackers' use of AI.Nov 2, 2026NCSC warns of AI-powered attacks on UK critical infrastructure by 2028security leaderThreat models must account for AI-accelerated, automated attack chains, and discovery and patching cycles must be adjusted accordingly.security operationsAs time-to-exploit shortens, automating patching and detection processes moves to the fore.public-sector decision-makerThe agency's assessment supports the case for urgency on critical infrastructure resilience and AI security regulation.Nov 13, 2026FBI: hackers use ARTEX AI agents to gain privileged access in under three hourssecurity leaderFaster acquisition of privileged access shortens the time defenders have to react.security operationsMonitoring of privileged accounts and rapid isolation should be strengthened.Nov 16, 2026FBI: attackers use open-source AI pentest tool ARTEX to breach networks in under three hourssecurity leaderAttackers' time to breach has shrunk to hours, so response times in defence plans need to be adjusted accordingly.security operationsEarly detection of privilege escalation and lateral movement is critical, because the tool reaches privileged access quickly.developerAs attackers are using legitimate testing tools, it is worth running AI-agent attack simulations in your own environment.Nov 16, 2026Google: Russian-speaking actors behind first fully autonomous AI cyberattacksecurity leaderGoogle says full automation has already appeared, so a human-speed attacker can no longer be the default assumption in threat models.security operationsAutomated attacks require faster detection and response.journalist and researcherThe claim of a first automated attack calls for independent verification and further technical detail.Nov 16, 2026Google: attackers use AI platforms and agents, fully autonomous attacks could followsecurity leaderThe threat model and defence plan must account for the speed of attacks automated with agents.security operationsThe agent-based automation observed in attacks means faster reconnaissance and intrusion.journalist and researcherGoogle is one source on the agent-based automation observed in live attacks.Nov 16, 2026Google: attackers use AI agents to turn software patches into working exploitssecurity leaderPatch-management SLAs should be re-evaluated in light of automated exploit development.security operationsThe window between a patch release and an exploit appearing is narrowing, so patching speed matters even more.developerPublished security fixes could be quickly reverse-engineered and exploited by AI agents.

Contested autonomy and excessive fears

Voices: VulnCheck, iTnews, Sysdig-kritikusok

Feb 2, 2026Reports differ on how autonomous AI-driven cyberattacks really aresecurity leaderThreat models need to account for the assisting role of agents; the degree of autonomy is disputed.security operationsPrepare for AI-accelerated attack steps by shortening detection and response times.journalist and researcherThe degree of autonomy should be examined separately from assistive use.May 20, 2026iTnews: fears of unfettered hacking over Anthropic's Mythos model likely overstatedsecurity leaderRisk assessments should weigh AI's offensive capabilities soberly rather than treat them with panic.journalist and researcherJudging the fears requires evidence of the models' actual offensive capabilities.Jul 6, 2026Sysdig: AI agent carries out ransomware attack, but 'no human help' claim is disputedsecurity leaderThe researchers say the attack chain runs at machine speed, so defence plans built on human reaction time need to be rethought.security operationsBecause an instance exploiting Langflow CVE-2025-3248 was involved, exposed Langflow deployments should be checked.developerFor systems built on Langflow, the concrete tasks are patching CVE-2025-3248 and assessing how exposed instances are.journalist and researcherThe "fully autonomous" label is disputed, so Sysdig's claim should be reported together with a precise description of the human role.Jul 29, 2026VulnCheck: only 1% of AI-discovered vulnerabilities exploited in the wildsecurity leaderAI-driven vulnerability research may increase the patching workload, while the exploitation rate remains low for now.journalist and researcherThe measured rate adds data to the debate on AI's offensive capabilities.Sep 15, 2026Taiwan reports first AI-assisted cyberattack from overseassecurity leaderAI-assisted attacks pose a faster, more automated threat.public-sector decision-makerThe official announcement indicates that state-level detection of AI-based attacks is under way.

Who is affected and how

security leader
Expect intrusion times in the order of hours: according to an FBI warning, attackers using ARTEX AI agents can obtain elevated privileges within three hours. Detection and isolation SLAs should be aligned with this, not with the 'first' label of campaign news.
security operations
Agent attacks occur at machine speed: Sysdig reports a 31-second attack, while another report details 2,250 Microsoft 365 tenants compromised within hours. Alerts should be built for rapid, mass authentication and reconnaissance sequences, and the SOC should be tested on its response time.
public-sector decision-maker
Based on government cases in Mexico, Taiwan, and Thailand, administrative systems and accounts are high-priority targets. Enforcing MFA and access control is more important than AI-specific regulation.
journalist and researcher
For every claim of a 'first autonomous attack', ask who determined autonomy based on what evidence and whether there is independent confirmation.

Hungary

No or minimal domestic trace during the period under review: the list contains no [HU] story. It is worth monitoring whether domestic incidents are retroactively labeled with AI tags without evidence.

Weekly trend

Supporting stories per week

2Nov 17–23: 1Nov 24–30: 1dec. 1–7.: 0dec. 8–14.: 0dec. 15–21.: 0dec. 22–28.: 0Dec 29 – Jan 4: 0jan. 5–11.: 0jan. 12–18.: 0jan. 19–25.: 1Jan 26 – Feb 1: 0febr. 2–8.: 1febr. 9–15.: 1febr. 16–22.: 1Feb 23 – Mar 1: 1Mar 2–8: 2Mar 9–15: 0Mar 16–22: 1Mar 23–29: 0Mar 30 – Apr 5: 0Apr 6–12: 0Apr 13–19: 0Apr 20–26: 0Apr 27 – May 3: 0May 4–10: 0May 11–17: 1May 18–24: 0May 25–31: 0Jun 1–7: 0Jun 8–14: 0Jun 15–21: 0Jun 22–28: 0Jun 29 – Jul 5: 1Jul 6–12: 1Jul 13–19: 0Jul 20–26: 0Jul 27 – Aug 2: 1Aug 3–9: 1Aug 10–16: 1Aug 17–23: 0Aug 24–30: 0Aug 31 – Sep 6: 0Sep 7–13: 2Sep 14–20: 0Sep 21–27: 0Sep 28 – Oct 4: 0Oct 5–11: 2Oct 12–18: 0Oct 19–25: 1Oct 26 – Nov 1: 0Nov 2–8: 0Nov 9–15: 1Nov 16–22: 2Nov 17–23Nov 16–22

Turning points

Related narratives in the News magazine

Supporting stories

Nov 17, 2025Anthropic says China-linked actors used AI agents in cyber espionage campaignsecurity leaderDefences must match the pace of an AI-accelerated, partly autonomous attack chain.security operationsAutomating detection and response is increasingly important against machine-speed attacks.public-sector decision-makerGovernment bodies are also among the targets mentioned in the report.journalist and researcherThe case is being analysed as the first AI-driven campaign presented as large-scale.Nov 24, 2025Anthropic discloses AI-orchestrated cyber-espionage campaign; Congress calls CEO to testifysecurity leaderThey must allow for the possibility that individual steps of an attack can be carried out by an AI agent, and adapt their defence model and detection accordingly.public-sector decision-makerThe congressional hearing signals that AI-driven espionage has become a question of regulation and liability.journalist and researcherAnthropic's account is the first documented case of AI-orchestrated attacks; its details still need to be verified from an independent source.Jan 19, 2026VoidLink cloud malware shows signs of being AI-generatedsecurity leaderAI is speeding up malware development, so detection assumptions for cloud environments should be reviewed.security operationsBehaviour-based detection becomes more important against fast-changing malware that targets the cloud.journalist and researcherThe markers that identify AI-generated code are a research area in their own right.Feb 2, 2026Reports differ on how autonomous AI-driven cyberattacks really aresecurity leaderThreat models need to account for the assisting role of agents; the degree of autonomy is disputed.security operationsPrepare for AI-accelerated attack steps by shortening detection and response times.journalist and researcherThe degree of autonomy should be examined separately from assistive use.Feb 12, 2026Google: state-backed groups used Gemini AI for reconnaissance and attack supportsecurity leaderAttackers are building generative AI into reconnaissance and attack planning, and the threat model needs to account for this.journalist and researcherGoogle's statement is a documented, vendor-confirmed example of AI misuse.Feb 20, 2026AWS says AI-augmented threat actor accesses FortiGate devices at scalesecurity leaderAI increases attackers' ability to scale, so rapid patching and reduced exposure are key to perimeter defence.security operationsFortiGate configuration, exposure and authentication should be reviewed urgently.Feb 23, 2026Amazon says Russian-speaking hackers used generative AI to compromise 600 firewallssecurity leaderAI can also increase the scale of attacks by less skilled attackers.security operationsAccess controls and configuration of perimeter firewalls need to be rechecked.Mar 4, 2026Hackers used Claude Code to build attack tools and steal data from Mexican government agenciessecurity leaderBecause coding agents are being used to develop attacks, abuse detection and pattern monitoring for agents are important.public-sector decision-makerAn AI-assisted attack on government data sets also raises data protection risks.Mar 7, 2026Microsoft says attackers use AI at every stage of cyberattackssecurity leaderThreat models should account for AI-assisted attacker activity at every stage of the chain.journalist and researcherMicrosoft's claim gives a starting point for documenting and independently verifying attackers' use of AI.Mar 16, 2026IBM links suspected AI-generated "Slopoly" malware to Hive0163 ransomware attackssecurity leaderAttackers can build malware faster and more cheaply with AI, which may challenge current detection assumptions.security operationsDetection should also rely on behavioural signals, not only on known signatures.journalist and researcherIBM's claim needs independent verification; samples and related details can be examined by other researchers.May 11, 2026Google detects AI-developed zero-day exploit tied to cybercriminal groupsecurity leaderAI-accelerated vulnerability research and exploit development mean shorter defence windows, which should be factored into threat models and patch priorities.security operationsBecause of the reported bypass of two-factor protection, it is worth reviewing how authentication processes are monitored.developerBecause AI can uncover flaws, code auditing and rapid patching matter even more.journalist and researcherThe case is a defender-documented example of AI-assisted exploit development that researchers should follow.Jul 2, 2026Researchers say AI agent ran ransomware attack on database via Langflow flaw without human involvementsecurity leaderAutomating the attack chain can shorten the time between discovery and extortion.security operationsLangflow instances need to be patched and isolated, and databases need strict access controls and backups.developerExposed instances of AI workflow frameworks are entry points for attackers.journalist and researcherThe researchers say the attack ran without human direction; no independent source has yet confirmed this.Jul 6, 2026Sysdig: AI agent carries out ransomware attack, but 'no human help' claim is disputedsecurity leaderThe researchers say the attack chain runs at machine speed, so defence plans built on human reaction time need to be rethought.security operationsBecause an instance exploiting Langflow CVE-2025-3248 was involved, exposed Langflow deployments should be checked.developerFor systems built on Langflow, the concrete tasks are patching CVE-2025-3248 and assessing how exposed instances are.journalist and researcherThe "fully autonomous" label is disputed, so Sysdig's claim should be reported together with a precise description of the human role.Jul 27, 2026Hackers used an autonomous AI agent to spy on Thailand's finance ministrysecurity leaderAgent-based attacks prepare defenders for fast, partly automated attack chains.journalist and researcherThe case is a documented example of attackers using AI agents.Aug 3, 2026China-based hacker uses DeepSeek agent against a security firm's networksecurity leaderThe pace and method of attacks carried out by autonomous agents pose a new detection challenge.security operationsLook for signs of proxyjacking, such as unusual outbound traffic and relay services.journalist and researcherThe attackers' use of an AI agent provides a documented example for the topic.Aug 11, 2026China-linked hackers attack Taiwan's government systems with autonomous AI agentssecurity leaderBecause of autonomous, open-source agent chains, the threat model must account for the speed of reconnaissance and account compromise.security operationsAnomaly monitoring of government accounts and protection of credentials come to the fore.public-sector decision-makerThe campaign shows that autonomous AI tools can now be used against state systems.journalist and researcherThe technical details of the open-source autonomous attack chain need analysis and verification.Sep 10, 2026Anthropic: AI lets smaller actors run state-level hacking campaignssecurity leaderThe report says AI appears in attacks at varying levels of autonomy, which affects threat modelling.security operationsAnthropic says attacker sophistication no longer depends on the size of the actor, so detection assumptions need rethinking.journalist and researcherThe report gives details of cases ranging from cyberattacks to influence operations, surveillance, biology and weapons building.Sep 10, 2026Hundreds of AI agents compromise PaperCut servers, GreyNoise sayssecurity leaderAttacks automated with AI agents speed up exploitation relative to the patching window.security operationsPatching the PaperCut vulnerabilities is urgent, as the campaign uses agents to scan and attack at scale.Oct 5, 2026Autonomous AI agent chains two Zammad zero-days to breach Dutch vulnerability disclosure institute DIVDsecurity leaderDefence plans should also account for the speed of automated, agent-based attacks.security operationsUpdating Zammad instances and reviewing the exposure of helpdesk systems may be urgent.developerAgents can chain vulnerabilities together at machine speed.Oct 8, 2026CrowdStrike suspects 26-year-old in China used Claude Code against South Korean bankssecurity leaderThe attribution rests on a single vendor's claim and is disputed by those involved, so it should be treated with caution.developerGiven the misuse of coding agents, provider logging and abuse detection are key issues.journalist and researcherUsing AI provider logs as attribution evidence raises new methodological questions.Oct 22, 2026China-linked AI agent breached 2,250 Microsoft 365 tenants within hours, bypassing MFAsecurity leaderMFA alone did not protect in this scenario, so tenant configurations should be reviewed.security operationsMicrosoft 365 tenant logs and access rights should be reviewed.Nov 13, 2026FBI: hackers use ARTEX AI agents to gain privileged access in under three hourssecurity leaderFaster acquisition of privileged access shortens the time defenders have to react.security operationsMonitoring of privileged accounts and rapid isolation should be strengthened.Nov 16, 2026Google: Russian-speaking actors behind first fully autonomous AI cyberattacksecurity leaderGoogle says full automation has already appeared, so a human-speed attacker can no longer be the default assumption in threat models.security operationsAutomated attacks require faster detection and response.journalist and researcherThe claim of a first automated attack calls for independent verification and further technical detail.Nov 16, 2026Google: attackers use AI platforms and agents, fully autonomous attacks could followsecurity leaderThe threat model and defence plan must account for the speed of attacks automated with agents.security operationsThe agent-based automation observed in attacks means faster reconnaissance and intrusion.journalist and researcherGoogle is one source on the agent-based automation observed in live attacks.

Machine analysis based on story titles and summaries, with a verification pass. The source of the facts is the article linked at each story.

What do you make of this piece?

The split shows after 20 votes.

Comment on LinkedIn
Saved pieces

    Weekly email

    One email a week with the main stories from the Agent Age editions. You confirm by email and can unsubscribe in one click.

    Almost done: we sent you an email. Open it and press the confirm link. If it does not arrive within a few minutes, check your spam folder.

    ↑