First production system to independently form cybersecurity concepts without human-defined rules
BRACEBRIDGE, ON / ACCESS Newswire / October 20, 2025 / SKADI Cyber Defense Corporation, a Canadian cybersecurity company, today announced the production deployment of the world's first autonomous ontology-based security AI. The system independently learns how cyberattacks work by observing them in live environments, eliminating the need for human-defined rules, labels, or taxonomies.
Unlike traditional security AI that reasons through frameworks built by experts, SKADI's system constructs its own conceptual understanding of how attacks unfold. By doing so, it addresses one of cybersecurity's core challenges: threats evolve faster than humans can encode detection logic.
"Every security system on the market today operates within human-designed taxonomies," said Rachel Clark, Founder and CEO of SKADI. "Experts tell the AI what reconnaissance means, what persistence looks like, how attack chains are structured. We've removed that dependency. Our AI observes security events and learns those relationships on its own, then continuously refines its understanding as threats evolve."
How Ontological AI Works
An ontology defines how concepts relate to one another. A child who learns that "mammal" includes "dog," which includes "golden retriever," is building an ontology through observation, not memorization.
SKADI's system does the same for cybersecurity. When it repeatedly observes process discovery followed by registry modification in a defined time window, it doesn't simply correlate the two. It forms a higher-order concept: reconnaissance precedes persistence. It builds that concept through self-supervised learning, tests it against new data, and refines its understanding over time.
Where existing AI systems reason using human-created ontologies like MITRE ATT&CK, SKADI's system autonomously generates its own. This enables it to detect novel or evolving attack chains without waiting for taxonomy updates from human analysts.
The Technical Distinction
Conventional ontology-based security relies on predefined structures: experts define relationships, and AI systems reason within those boundaries. SKADI's approach is different. It induces concepts and relationships directly from observed event sequences using unsupervised learning and temporal pattern induction.
When a new behavior pattern emerges-say, a persistence method unseen in existing frameworks-the system identifies it, integrates it into its conceptual model, and distributes that learning across customer environments automatically.
"This creates genuine network effects in enterprise defense," added Clark. "Every deployment contributes new insight, strengthening protection across the network. Traditional systems stay static until vendors push updates."
Canadian Innovation and Global Impact
The breakthrough represents a major milestone for Canadian technology innovation. While global competitors focus on automating rule creation, SKADI's approach eliminates the rule layer entirely, allowing the AI to construct its own conceptual framework of security knowledge.
The system's ability to learn directly from real-world behavior allows organizations to scale security operations at radically lower cost, without sacrificing adaptability or precision.
"This kind of AI wasn't technically achievable 18 months ago," said Clark. "Language models only recently reached the capability threshold required for autonomous concept formation. We're first to bring it into production, and we built it entirely in Canada."
In the long run, this approach will change how cybersecurity scales. Organizations will no longer need vast teams of analysts to keep rule sets and detection logic current. The AI's understanding evolves automatically, reducing operational overhead while improving speed and accuracy. As deployments grow, each environment strengthens the collective intelligence of the system: creating shared protection that improves for everyone, everywhere it's installed.
For more information about "Frostbow" and SKADI's full suite of cybersecurity services, visit skadicyber.com.
Contact Information
Shantelle Dedicke
Frances Roy Agency
shantelle@francesroy.com
3076317644
Rachel Clark
Chief Executive Officer/SKADI
rachel.clark@skadicyber.com
SOURCE: SKADI Cyber Defense
View the original press release on ACCESS Newswire:
https://www.accessnewswire.com/newsroom/en/computers-technology-and-internet/skadi-cyber-defense-corporation-deploys-autonomous-self-learning-1089379
