Zero-Trust. Extended.
The potential impact of AI on cybersecurity is immense and growing, notable examples being:
Offence: Speed of attack is an order of magnitude higher than humans can respond to.
Offence: The speed, breadth, and depth of AI-powered attacks create an explosion in false positives, overwhelming already stretched teams and obscuring stealth threats.
Defence: AI-powered, automated defence architectures are needed as a response to this.
AI means the democratisation of sophisticated cyber-attacks; it is no longer the exclusive realm of nation-state actors. Sophisticated, multi-stage stealth attacks are within reach of many actors.
AI can bring huge benefits to cybersecurity tools, but as the emphasis on speed and volume of inference grows, a small, vital (and until now, overlooked) step in the process will greatly impact the ability of AI to reach its potential in this space. That step is to answer this question:
“Is the telemetry the AI is acting on true or false?”
This is fundamentally important: if AI reasons over falsified or wrong data, then users simply end up in a worse place, even faster; this will limit the acceptance of AI in cyber-defence. The reasons for false data can be intentional (e.g., cyber-attacks) or unintentional (e.g., tool misconfigurations), but the credibility of AI to add value in cybersecurity depends on the answer. PILOT is the answer.
PILOT = Parallel Integrity Layer of Truth.
Most cybersecurity tools infer meaning from the telemetry available to them.
PILOT infers meaning about the telemetry it sees.
Rather than assuming reported data accurately reflects what has occurred, PILOT independently tests whether the reported system state is physically and logically possible.
PILOT is a patent-pending integrity layer designed to work alongside existing cybersecurity tools, helping make the data they rely on more trustworthy without requiring rip and replace.
PILOT verifies whether reported activity is physically and logically credible across IT, OT and converged environments. It does not add another anomaly detector; instead, it provides a definite verdict, valid, violation or uncertain, on whether the whole picture holds together, using one verification layer across single or multiple cyber domains.
Independent, EU-funded and peer-reviewed research has demonstrated the feasibility of our core approach on industrial testbeds. Published in December 2025, after our patent priority dates, the research was conducted entirely independently and without any involvement from Aequus. Aequus enables AI-driven cyber defence to operate on more trustworthy data, improving outcomes without requiring existing tools to be replaced.
How PILOT works:
