📊 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 shows AI is enabling cybercriminals to perform more complex and dangerous attacks. Traditional threat metrics no longer reliably distinguish high-risk actors, as AI levels the playing field. This shift has significant implications for cybersecurity defenses.
New research from Anthropic indicates that AI is fundamentally changing the landscape of cyber threats, making attackers more capable and harder to distinguish using traditional metrics. The report analyzed 832 banned accounts involved in malicious activities over a year, revealing that AI tools are now enabling even less skilled actors to perform complex, high-risk actions within compromised networks. This development challenges long-standing threat assessment models, which relied on the number of techniques used and the sophistication of tools to gauge danger.
Anthropic examined 832 accounts flagged for malicious cyber activity between March 2025 and March 2026, mapping their tactics onto the MITRE ATT&CK framework. The findings show that 67.3% of these accounts used AI to prepare for attacks, primarily for malware creation. More notably, 6.5% employed AI for advanced activities like lateral movement within networks, with this proportion increasing from 33% to 56% over the year. The shift toward post-infiltration activities highlights that AI is enabling broader segments of threat actors to perform complex operations previously limited to highly skilled hackers.
Furthermore, the report notes that the traditional markers of threat level—such as the number of techniques or the tools used—no longer reliably distinguish high-risk actors. Both novice and expert actors now employ similar technique counts and tools, as AI supplies many of the technical capabilities. Instead, the primary differentiator appears to be where in the attack lifecycle they deploy AI: more dangerous actors focus on operationally demanding tasks like account discovery and lateral movement, but even this signal is weakening as AI democratizes advanced capabilities.
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
Artificial Intelligence for Cybersecurity: Develop AI approaches to solve cybersecurity problems in your organization
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.
Ubiquiti Unifi Security Appliance (USG), Single,White
Integration with Unifi Controller. Powerful firewall performance
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.
Evasive Malware: A Field Guide to Detecting, Analyzing, and Defeating Advanced Threats
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.
AI-Enabled Threat Detection and Security Analysis for Industrial IoT
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 of AI’s Role in Modern Cyber Threats
This shift means cybersecurity defenses based on traditional threat metrics are increasingly ineffective. As AI enables less skilled actors to perform sophisticated operations, the threat landscape becomes more unpredictable and difficult to assess. This democratization of attack capabilities raises the risk of widespread, high-impact breaches by actors with minimal technical expertise, complicating efforts to prioritize threats and allocate resources effectively.
Organizations and security professionals must reconsider their threat models, focusing less on technique counts and more on the context and timing of AI deployment within an attack. The evolving threat landscape underscores the need for adaptive detection methods and more sophisticated threat intelligence that can account for AI-enabled attack behaviors.
Evolution of Cyberattack Techniques in the AI Era
For decades, threat assessment relied on quantifying attacker techniques and tool sophistication, with more techniques indicating higher danger. The MITRE ATT&CK framework provided a standardized way to categorize tactics, supporting this approach. However, recent developments show that AI tools are now automating complex tasks, reducing the importance of technical skill and tool variety as indicators of threat level. The rise of AI-assisted lateral movement and account discovery marks a significant shift in attack strategies, reflecting broader trends of AI integration into cybercrime since early 2025.
This trend aligns with earlier warnings from cybersecurity experts about AI’s potential to lower barriers for malicious activity, but the Anthropic report offers concrete data illustrating how these changes are unfolding in real-world scenarios over the past year.
“The traditional methods of threat assessment are no longer sufficient. AI is leveling the playing field, making even less skilled actors capable of high-impact attacks.”
— Thorsten Meyer, AI security researcher
Unclear Impact of Evolving Threat Detection Methods
It remains uncertain how current cybersecurity defenses will adapt to these changes. While the report highlights the increasing difficulty of threat differentiation, specific strategies for countering AI-enabled attacks are still emerging. Additionally, the full scope of how widespread AI use is among malicious actors remains unknown, as the analysis is based on a subset of flagged accounts.
Future Directions in Cybersecurity and AI Threats
Cybersecurity professionals will need to develop new detection frameworks that focus on attack context, timing, and behavioral patterns rather than technique counts alone. Expect increased investment in AI-driven threat intelligence and adaptive defense systems. Monitoring how threat actors continue to leverage AI will be critical, with ongoing research and real-time data collection shaping future security strategies.
Key Questions
How does AI make attackers more dangerous?
AI automates complex attack tasks, such as lateral movement and account discovery, enabling less skilled actors to perform high-impact operations inside networks.
Why can’t traditional threat metrics detect these new threats?
Because AI supplies many of the technical capabilities, the number of techniques or tools used no longer correlates with threat level, making old heuristics ineffective.
What should organizations do to improve security?
Organizations need to adopt more behavioral and context-aware detection methods, focusing on attack timing and operational patterns rather than technique counts alone.
Is this trend likely to continue?
Yes, as AI technology becomes more accessible and sophisticated, its integration into cyberattack strategies is expected to grow, further complicating threat detection.
Source: ThorstenMeyerAI.com
