Penta Security, Autocrypt Publish AI-Era Cyber Security Declaration
Eight experts from the two firms set out three cyber defence principles and propose external controls for AI behaviour.
Penta Security and Autocrypt have jointly published a "Cyber Security Declaration" setting out basic cybersecurity principles for the AI era and directions for safe AI implementation, dailysecu.com reported. The declaration is dated October 27, 2026.
Eight experts from the two companies took part, including Autocrypt's chairman and CEO and Penta Security's CEO, covering cryptography, software engineering, information security and physical AI.
At the centre of the document are three "Cyber Defence 3 Principles": Observability, Cross-Layer Defence and Zero Damage Design.
The declaration also proposes external control measures for AI, including verifying and controlling AI behaviour and clarifying who is responsible in cases where AI safety rules are bypassed.
Quick answers
What are the Cyber Defence 3 Principles in the declaration?
Observability, Cross-Layer Defence and Zero Damage Design.
Which companies published the Cyber Security Declaration?
Penta Security and Autocrypt, with eight experts from the two firms taking part.
What does the declaration propose for controlling AI?
External control measures, including verifying and controlling AI behaviour and clarifying responsibility when AI safety rules are bypassed.