Loading story…
Palo Alto Networks Unit 42 warned that frontier and commercially available agentic AI is shifting offensive capability toward attackers, enabling lower-skilled actors to conduct vulnerability discovery, reverse engineering, malware development, social engineering, and other attack-chain activities at machine speed. In an investigation involving a customer, Unit 42 said an attacker reportedly used an agentic framework to exploit 50 applications and other enterprise weaknesses in under 10 hours, including exploitation, privilege escalation, and data theft.
Unit 42 said AI is already supporting malware creation, task delegation, ransomware negotiations, and reconnaissance, while making attribution more difficult. It identified identity compromise, software supply-chain exposure, and nation-state discovery of enterprise weaknesses as key risks, and urged organizations to adopt zero-trust controls, continuously train employees, govern internal AI use, and build adaptable security strategies for agentic AI.
Track how attackers are adapting to this technology.
7 events from the most recent confirmed update back to the earliest known activity.
Unit 42 reported an AI-enabled operation that identified and exploited vulnerabilities, escalated privileges, and stole data within 10 hours, work it said would normally take a penetration-testing team about two weeks.
Unit 42 began investigating an attack against one of its customers in which an attacker reportedly used an agentic framework to exploit 50 applications and other enterprise weaknesses in under 10 hours.
During three weeks of internal testing with AI penetration-testing models, Unit 42 said it completed work comparable to one or two years of conventional penetration testing and uncovered dozens of vulnerabilities across customer environments.
Five months after its estimate, Unit 42 reported seeing early waves of the anticipated AI-enabled threat activity in the wild.
Through its Project Glasswing evaluations, Unit 42 estimated that the assessed AI capabilities could become available to attackers within one year.
Palo Alto Networks Unit 42 launched Frontier AI Defense, combining threat intelligence, threat telemetry, and frontier AI models for enterprise security use cases.
Anthropic formed Project Glasswing with Palo Alto Networks and other technology companies to identify and remediate security defects using Anthropic's Mythos model.
Follow how adversaries are adapting to this technology, and where it touches your stack today.
2 references tracked. Mallory keeps watching after this page renders.
Map indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
Security leaders are warning that **agentic AI** is shifting the threat model by turning authenticated, permissioned access into a primary attack path. Symantec argues that controls designed around human behavior and pace (e.g., static credentials, overprivileged access, and human-scale monitoring) are poorly suited to autonomous agents that can operate at machine speed, rapidly probing APIs and systems and using existing permissions to move laterally. The post highlights proof-of-concept and prior incidents where agent behavior can be manipulated to expose secrets (e.g., coercing a voice agent to reveal an API key), framing an agent’s own decision logic as an exploitable attack surface that drives the need for redesigned authorization and tighter identity/access models. Microsoft threat intelligence reporting, echoed in an interview with *The Register*, describes threat actors using AI agents to automate “janitorial” tasks that materially increase operational tempo—especially **reconnaissance** and **attack infrastructure** setup/management. Sherrod DeGrippo (Microsoft) said autonomous agents can be tasked to scan target netblocks, inventory compromised environments, and accelerate standing up infrastructure, and Microsoft has observed North Korea-linked **Coral Sleet** leveraging development platforms to quickly create and manage infrastructure supporting campaigns (including activity associated with the fake IT worker scam). Together, the reporting indicates agentic AI is already being applied to make attacks **better, bigger, and faster**, with defenders needing to account for automated, high-velocity activity that may originate from valid credentials and legitimate tooling.
Jun 29, 2026
Security researchers reported that threat actors are using **agentic AI** and widely available large language models to dramatically speed up cloud intrusions, turning familiar attack paths into high-tempo operations. In one documented case cited by Sygnia, a lone attacker compromised an **AWS** environment in about **72 hours** after stealing an access key from an internet-facing application, then used AI-assisted workflows to harvest secrets, establish persistence, exfiltrate data, and carry out disruptive actions consistent with extortion. The intrusion did not depend on zero-days or novel malware; instead, it amplified standard cloud attack techniques through automation, concurrency, and faster decision-making. Broader industry analysis from Sysdig and Google indicates that AI is not fundamentally changing initial access methods so much as increasing the **speed, scale, orchestration, and evasion** of post-compromise activity. Researchers said the most exposed organizations are those carrying longstanding security debt, including weak secrets management, excessive permissions, poor identity governance, limited monitoring, unpatched systems, and inadequate control over AI-enabled assets. Recommended defenses include tightening repository and credential access, enforcing network segmentation and outbound restrictions, routing traffic through **WAFs**, improving visibility and incident readiness, inventorying AI agents with production access, and treating AI systems as monitored production assets.
Aug 12, 2026
Palo Alto Networks' Unit 42 reported that artificial intelligence is currently serving mainly as a force multiplier for attackers, speeding up and scaling established intrusion workflows rather than introducing entirely new attack methods. Drawing on incident response investigations, the report said threat actors still rely on familiar tactics including **credential theft**, **phishing**, exploitation of known vulnerabilities, and **ransomware**, while using AI to accelerate malware development, automate content generation, and streamline reconnaissance. Unit 42 leaders Andy Piazza and Richard Emerson said AI-enabled attacks have not yet required enterprises to redesign core defense strategies, but warned that broader attacker adoption could strain detect-and-respond operations and raise the value of prevention-focused controls. The report also pointed to emerging activity including **AI-assisted malware**, malware that queries **LLM** or **MCP** servers for command-and-control guidance, **agentic ransomware**, and **token jacking** aimed at cloud AI services and LLM API credentials.
Jul 17, 2026