Outlook

Three transitions shape the near future: the automation of combat by AI, the legal codification of security and the cryptographic overhaul.

On the attack side: perfect phishing, industrialised deepfakes, automated vulnerability research. On the defence side: behavioural detection, augmented SOCs, automated response. A third front: securing AI systems themselves: prompt injection, data poisoning, agent governance.

A sufficiently powerful quantum computer would break part of today's asymmetric cryptography. The response is under way: in 2024, NIST published the first post-quantum cryptography standards. Organisations must inventory their uses of cryptography before migrating.

Europe is imposing a de facto global standard: NIS2, DORA for finance, and the Cyber Resilience Act for connected products. Cybersecurity is no longer good practice: it is becoming a legal obligation, and a market.

Passkeys are gradually replacing the password with a cryptographic key tied to the device and to biometrics, much less vulnerable to phishing. Combined with zero trust, they are quietly overhauling authentication worldwide.

Cyberspace is settling in as a domain of confrontation below the threshold of war: massive espionage, pre-positioning, targeted sabotage, information operations. International regulation is progressing slowly; resilience is becoming the realistic goal.

Whatever the tools, the last line of defence remains culture: trained users, reflexes of verification, organisations that rehearse for crises. The cybersecurity of 2030 will be more automated, but no less human.