Know what is on your site before a plaintiff’s firm does.Free website scanScan Your Site
Log in Sign up Book a demo
Home› Editorial› Spain’s CCN Issues ENS Guides on Offensive AI…
GOVERNANCE RISK AND COMPLIANCE

Spain’s CCN Issues ENS Guides on Offensive AI and the Quantum Threat: What CCN-STIC 861 and 862 Ask of Organisations

Spain’s National Cryptologic Centre (CCN), part of the National Intelligence Centre, has published two new guides on applying the country’s National Security Scheme, the Esquema Nacional de Seguridad (ENS), to two threats that worry security teams most: attackers using artificial intelligence and quantum computers that.

Nov 20, 2026 8 min read

Know what’s on your site before a plaintiff’s firm does

A free privacy audit shows which cookies, pixels and trackers are running — and which ones commonly appear in privacy claims.

Get My Free Audit

Spain’s National Cryptologic Centre (CCN), part of the National Intelligence Centre, has published two new guides on applying the country’s National Security Scheme, the Esquema Nacional de Seguridad (ENS), to two threats that worry security teams most: attackers using artificial intelligence and quantum computers that could eventually break today’s public-key cryptography. Two days later it added a national roadmap for moving to post-quantum cryptography, with dates running to 2035. None of it changes the ENS itself. What it changes is how auditors and security officers are expected to read the measures they already have, and that matters for every company that runs systems for Spain’s public sector.

TL;DR

  • The CCN published CCN-STIC 861 on offensive AI and CCN-STIC 862 on the quantum threat in November 2026, giving criteria for 25 and 26 ENS Annex II measures respectively, each with a matrix of 11 risk scenarios.
  • Neither guide changes the ENS catalogue of measures or which measures are mandatory; both work through the existing risk analysis and Statement of Applicability.
  • A separate CCN roadmap published November 19 sets May 31, 2027 for governance, policies and an initial cryptographic inventory, and the end of 2030 for migrating high-risk systems.

What was published

According to the CCN’s announcement of November 17, 2026, the two documents join the CCN-STIC 800 series, the guides that accompany the ENS:

  • CCN-STIC 861, Criteria for the applicability of ENS measures against offensive AI (21 pages).
  • CCN-STIC 862, Criteria for the applicability of ENS measures against the quantum threat (22 pages).

The ENS is set out in Royal Decree 311/2022. It applies to public sector information systems and, under Article 2.3, to private companies that provide services to public bodies under contract. Any other organisation can adopt it voluntarily. The second additional provision of the decree gives the CCN the job of producing guides to help with compliance, and both new documents describe themselves as technical guidance within that role.

Some English summaries describe a single set of “guidelines” on both the “National Security Scheme” and the “National Security Framework.” Those are two translations of the same thing, the ENS, and the publication is two separate guides with different scopes. The often-quoted language about preparing, planning, executing and monitoring the post-quantum transition “in a homogeneous, proportionate, verifiable and risk-based manner” is the stated purpose of CCN-STIC 862 specifically.

The shared idea: same controls, read differently

Both guides start from the same position. The ENS is a valid framework for these threats, and the right response is to adapt and strengthen existing measures rather than to invent new ones. Each guide says it does not modify the catalogue of measures or their mandatory regime. Whether a measure applies still depends on the system’s category, its security dimensions and levels and, where relevant, the risk analysis. The guides tell organisations when an applicable measure has to take AI or quantum risk specifically into account.

The paperwork is also the same. Decisions on applicability, adaptation, priority and risk acceptance must be justified in the risk analysis and reflected, where appropriate, in the Statement of Applicability and risk treatment plans, with owners, deadlines, dependencies, exceptions, compensating measures and residual risk documented.

CCN-STIC 861: offensive AI as a multiplier

The AI guide does not treat offensive AI as a brand-new threat. It calls it a multiplier of capabilities and actors that makes known techniques such as reconnaissance, social engineering, vulnerability exploitation and malware generation faster, larger in scale, more personalised and more autonomous. It builds on an earlier CCN-CERT good-practice report, BP/36, and sets verifiable criteria for 25 Annex II measures.

One principle stands out: protection should not depend on proving that a particular attack used AI. What matters is whether the growth of those capabilities changes the likelihood, the impact, the time available to react or how well safeguards work. The guide also covers attacks on AI that an organisation has built into its own systems or supply chain, such as agent hijacking, prompt injection and abuse of delegated credentials. Broader AI governance duties, such as inventory, purpose, bias and human oversight, are explicitly left to a separate CCN document.

Its 11 risk scenarios are a useful checklist even outside Spain:

  • AI-generated phishing and impersonation
  • Automated discovery and exploitation of vulnerabilities
  • AI-generated or adapted malicious code
  • Prompt injection against AI agents
  • Manipulated AI outputs passed unchecked to other components or APIs
  • Poisoning of retrieval (RAG) knowledge sources
  • Abuse of permissions or excessive agency by an AI agent
  • Exfiltration of data, secrets or instructions through a model or agent
  • Supply chain compromise through models, data, libraries or APIs
  • Abusive consumption that exhausts capacity, quotas or budget
  • Coordinated multistage compromise that shrinks the time to detect and recover

Each is mapped to ENS measures, from authentication and segregation of duties to activity logging, application development and continuity testing. Exercises such as assured adversary simulation are one way to check whether those controls hold up. On priorities, the guide says actions should be ranked by the risk they reduce, not by how many there are, and that cost or difficulty alone does not justify leaving an unaccepted risk untreated.

CCN-STIC 862: preparing for the quantum transition

The quantum guide focuses on public-key mechanisms used for key establishment, authentication and digital signatures, and on the risk of encrypted data being captured now to be decrypted later. It argues that the answer is not just swapping algorithms but a gradual process of risk management that relies on four capabilities: a cryptographic inventory, an assessment of cryptographic and quantum risk, a transition plan, and crypto-agility, meaning the ability to replace algorithms, keys, certificates or libraries in a controlled way.

It sets criteria for 26 Annex II measures and maps 11 scenarios to them. These include capture now and decrypt later, broken authentication and signatures, unknown cryptographic dependencies, lack of crypto-agility, incomplete or inconsistent migration, vendor lock-in, performance impacts of larger post-quantum keys, interoperability failures and emergency migration. Priority should go to essential services, information that must stay confidential for a long time, public key infrastructure, identity and e-signature services, and assets exposed to capture now and decrypt later.

Several criteria will matter to anyone who signs contracts or manages records:

  • The asset inventory should record relevant cryptographic assets and a map of their dependencies, including purpose, version, location and supplier, without any secret material. A good data inventory is a useful starting point.
  • Contracts should include commitments on cryptographic maintenance and updates, vulnerability notification, certificate replacement, algorithm replacement, end of cryptographic support and cooperation in testing and migration, extending to subcontractors.
  • For electronic signatures, new signatures and the preservation of old ones are handled separately. The guide warns that late renewal or re-sealing cannot restore evidence whose authenticity can no longer be proven.
  • Standardisation is not authorisation. Only post-quantum or hybrid mechanisms authorised for the use case should be used, following CCN guidance.
  • Accepting the risk of vulnerable mechanisms must be explicit, reasoned, temporary and reviewable, with review at least once a year.

The roadmap’s dates

On November 19, the CCN published its roadmap for the transition to post-quantum cryptography in Spain, aimed at both the public and private sectors and coordinated with the ENS. It notes that a cryptographically relevant quantum computer could be available between 2030 and 2035 and calls for transition plans to start immediately. Its reference timetable:

  • By May 31, 2027: governance, policies and an initial inventory of cryptographic assets in place, use cases classified, and pilots for high- and medium-risk systems started where possible.
  • By December 31, 2030: migration of high-risk systems completed.
  • By December 31, 2035: migration of medium-risk systems completed, and low-risk systems where technically and economically feasible.

The CCN links this to the EU’s NIS Cooperation Group roadmap, which expects member states to take first steps, including a national roadmap and the start of planning and pilots, by the end of 2026. For teams buying technology, it pairs well with the post-quantum questions for vendors from CISA, the NSA and Australia’s ACSC.

What ENS-bound suppliers should do next

  1. Confirm your status. If you provide services to Spanish public bodies under contract, the ENS applies to those systems, and auditors are likely to read these guides as the expected interpretation.
  2. Add the scenarios to your risk analysis. Run the 11 AI scenarios and the 11 quantum scenarios against each system in scope and record the outcome, including where a scenario does not apply.
  3. Start the cryptographic inventory now. The May 2027 roadmap date is about six months away, and inventories of libraries, certificates, hardware security modules and embedded dependencies take time.
  4. Update supplier terms. Add cryptographic update and migration commitments, in line with your existing contractual safeguards, and push them down to subcontractors.
  5. Review AI agent permissions. Map which tools and credentials each agent can use, log its actions and validate its outputs before other systems act on them.
  6. Document risk acceptance. Where something cannot be fixed yet, write down why, for how long, and what compensating measure applies.

Spain is not the only country tightening its national cyber posture. Norway’s national security authority raised its cyber risk level to high and paired it with a list of practical measures. Spain’s approach is to work through documented, risk-based controls rather than new technology-specific rules.

FAQ

Are CCN-STIC 861 and 862 legally binding?

They are guidance issued under Royal Decree 311/2022. They do not change which ENS measures are mandatory, but they set out how the CCN expects those measures to be applied, so they will shape audits.

Do private companies have to follow them?

Those whose systems provide services to Spanish public bodies under contract fall within the ENS. Others can use the guides voluntarily as a reference.

Do the guides name specific post-quantum algorithms?

No. CCN-STIC 862 says specific mechanisms, parameters, authorisations and timelines will be set in separate CCN guidance.

Does CCN-STIC 861 cover AI governance?

No. It covers offensive use of AI and attacks on AI components. General AI governance obligations are left to a separate document.

It is a reference date in the CCN roadmap for high-risk systems, not a statutory deadline, but it signals what Spanish authorities will expect.

Spain’s message is that the controls already exist. What it wants is proof that organisations have checked them against these two threats before attackers do.

Put this into practice

Turn privacy guidance into working controls.

Captain Compliance helps teams discover tracking technologies, enforce consent, manage privacy requests and document the evidence behind every decision.

Book a demo Run a free scan