📊 Full opportunity report: The Frameworks Can’t See the Thing That Matters: A Year of AI-Enabled Cyber Threats on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
A year-long analysis reveals AI is empowering less skilled cybercriminals to conduct complex attacks, undermining traditional threat assessment frameworks. The shift toward AI-driven post-compromise techniques complicates detection and response efforts.
A new analysis from Anthropic shows that AI is fundamentally changing the landscape of cyber threats by enabling less skilled attackers to perform complex, post-intrusion activities, thereby undermining traditional threat assessment methods.
Anthropic examined 832 accounts banned for malicious activity between March 2025 and March 2026, mapping their techniques onto the MITRE ATT&CK framework. The findings reveal that AI is increasingly used to automate attack preparation, such as malware creation, with 67.3% of actors employing AI for this purpose. More concerning is the rise in AI-assisted lateral movement and account discovery activities, which increased significantly over the year, with the share of medium-risk or higher actors jumping from 33% to 56%.
Furthermore, the report indicates that AI’s role has shifted from initial access techniques like phishing toward deeper, post-compromise activities. These tasks, traditionally requiring high skill levels, are now performed by less sophisticated actors using AI, leading to a democratization of advanced attack capabilities. This development challenges the longstanding assumption that only highly skilled actors can carry out complex intrusions.
The frameworks can’t see the thing that matters
For decades, danger meant which techniques an attacker commands. A year of real AI-enabled attacks — 832 banned accounts mapped onto MITRE ATT&CK — shows that signal breaking, just as a new, harder-to-see one takes over.
A year of real misuse, mapped to the standard taxonomy
A window, not a census — these are the cases with enough detail to assess techniques thoroughly. Inside it, the risk level climbed fast.
WHAT WAS STUDIED
THE RISK CLIMB · MEDIUM-OR-HIGHER ACTORS
INTELLIGENT CYBERSECURITY SOFTWARE SYSTEMS: Threat detection automated response and adaptive defense architectures
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
“More techniques” stopped meaning “more dangerous”
The old heuristic: count the techniques, judge the tooling. AI dissolved it — because the model supplies the techniques either way. Watch the old signal fail, then watch what it misses.
Risk score vs. technique count
Two ways to read the same attacker. One is going blind. Press play.
WatchGuard Firebox T145 with 1 Year Basic Security Suite – Tabletop Firewall, 2.5Gb, 1Gb & SFP Ports, Enterprise Security for Branch Locations (WGT145031)
Watchguard T145 Firebox with 1 Year Basic Security Suite License (WGT145031) – The Firebox T145 delivers enterprise-grade protection…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Deeper into the attack — and into less-skilled hands
Across the year, AI use drifted from getting in toward acting once already inside — the operationally demanding stages that used to require an expert.
The attack lifecycle · where AI is now applied
The center of gravity moved right — toward post-compromise work.
Network Intrusion Detection
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From “what they know” to “what they’ve built”
The report sorts the signals into three tiers — one dead, one fading, one durable.
Technique count & tooling
16 vs. 20 between novice and expert; platform doesn’t correlate. The model supplies the techniques either way.
Where in the lifecycle AI is applied
Concentrating on operationally demanding, post-compromise stages is a better signal — but it’s eroding as the whole population heads there.
The scaffolding around the model
Architectures that let the model chain stages and run with minimal human input. Not what they know — whether they’ve built a system that lets AI run the attack.
Cyber Threat Hunters Handbook: Applying advanced analytics, automation, and collaborative intelligence for digital defense (English Edition)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Fixing the map before the territory moves again
A taxonomy that can’t name the most dangerous behavior on the field will quietly mislead the people relying on it. The response runs in two directions.
Fed back into the models
The findings informed safeguards on the most capable models, built to detect & block some of what was observed:
- Blocking malware development
- Blocking mass data exfiltration
- Putting tools in defenders’ hands first (Project Glasswing)
Taking it to the source
Following the Verizon work, Anthropic says it’s in discussions with MITRE about how ATT&CK might evolve:
- A vocabulary for agentic orchestration
- Naming the scaffolding that makes a model an operator
- An interactive technique visualization on the Red blog
Reading it in proportion
- The 832 cases are a detailed subset, not the full population — the precise percentages are directional, not definitive.
- “More autonomous” is not “fully autonomous” — even the standout case needed human input at key moments, which is itself a place for defenders to intervene.
- This is one vendor’s window — the company with visibility into misuse of its own model, publishing what it found. The right thing to do with the data, and worth remembering as you read it.
Implications for Threat Detection and Defense Strategies
This shift means that traditional threat assessment metrics—such as the number of techniques used or the sophistication of tools—are no longer reliable indicators of danger. The ability of AI to automate complex tasks blurs the distinction between skilled and unskilled attackers, making it harder for defenders to prioritize threats based on conventional heuristics. As post-compromise activities become more accessible, organizations face increased risks from less skilled actors capable of causing significant damage, necessitating new detection and mitigation approaches.
Evolution of Cyberattack Techniques in the AI Era
Historically, threat assessment relied heavily on counting techniques and analyzing tools used by attackers, assuming that more techniques indicated greater threat. However, recent developments show AI’s role in automating complex tasks, reducing the importance of technique diversity as a threat indicator. The report from Anthropic builds on prior concerns about AI’s influence on cybercrime, providing a real-world dataset that illustrates how AI is lowering the skill barrier and expanding attack capabilities across the board.
“The traditional markers of threat level—technique count and tooling—are no longer reliable indicators of danger in an AI-enabled threat landscape.”
— Anthropic research team
Unclear Impact of AI on Future Threat Detection
It remains uncertain how cybersecurity defenses will adapt to these changes. While the report details the current state, the evolution of attacker strategies and the development of new detection methods are ongoing. The extent to which AI will continue to democratize attack capabilities or lead to new defensive innovations is still unclear.
Next Steps in Cybersecurity Response and Research
Organizations will need to reevaluate threat assessment models, incorporating AI-driven indicators and focusing on post-compromise activities. Further research is expected to explore new detection techniques that can identify AI-assisted malicious behaviors and assess how threat actors will evolve in response to emerging defenses. Monitoring developments and updating security protocols will be critical in the coming months.
Key Questions
How does AI change the way we assess cyber threats?
AI automates complex attack activities, making it harder to distinguish between skilled and unskilled attackers based on techniques alone, thus undermining traditional threat metrics.
Are less skilled attackers now capable of performing advanced cyberattacks?
Yes, AI tools enable less skilled actors to carry out complex post-intrusion activities, such as lateral movement and account discovery, which previously required high expertise.
What can organizations do to improve detection of AI-enabled attacks?
Organizations should focus on monitoring post-compromise behaviors, integrating AI-aware detection systems, and updating threat models to account for AI-driven attack techniques.
Will current cybersecurity frameworks remain effective?
Likely not entirely; traditional metrics are less reliable, and new approaches will be necessary to identify and respond to AI-augmented threats effectively.
Source: ThorstenMeyerAI.com
