Albania’s Classified Information Amendments: A Structural Reading

The Tirana Examiner Legal Desk provides institutional analysis of Albanian legislation and judicial developments. This analysis is based on the draft amending text and Law 10/2023 as enacted.

Renada Bici  | Tirana Examiner Legal Desk

The draft amendments to Law 10/2023 on classified information arrived in the Albanian parliament without significant public commentary. That silence is itself a structural problem. Legislation governing the architecture of state secrecy — who may classify, who may declassify, who enforces, and under what conditions destruction is lawful — does not belong in the category of technical corrections. It belongs in the category of constitutional hygiene. These amendments deserve to be read as a coherent set of institutional choices, not merely as adjustments to existing procedure.

Read as a whole, the draft does three things simultaneously: it deepens an existing concentration of executive authority over the classification system; it creates new institutional actors and powers without fully closing the accountability loops around them; and it extends the legal reach of AKSIK, the national authority responsible for classified information, in ways that raise legitimate questions about due process. None of these moves is necessarily indefensible. But several require scrutiny that the legislative record does not appear to have supplied.

The Deepening of Executive Control Over Classification

The parent law already structures classification authority around the Prime Minister. Under Article 12 of Law 10/2023, the President and the Prime Minister are the only direct classifying authorities. Institutional heads acquire classification rights only by PM authorisation. The architecture of executive gatekeeping was built into the original law.

The draft amendments do not create this structure. They tighten it. The revised definition of “classifying authority” in Article 4 makes PM authorisation explicit in terms that the original definition handled more loosely. More significantly, the new Article 12.3 inserts a second-order exception: a senior official within a classifying institution may temporarily exercise classification rights, but only through a reasoned request from the institutional head, routed through AKSIK, subject to PM approval, within a defined time limit. AKSIK has seven calendar days to relay the request. The PM’s decision window is not specified.

The stated logic is coherent. Friction in the classification process should, in principle, discipline over-classification — the well-documented tendency of bureaucracies to protect information for reasons of institutional self-interest rather than genuine security need. A mandatory PM authorisation requirement imposes that friction.

The problem is not the mechanism. It is what the mechanism leaves unaddressed. Neither the parent law nor the draft establishes criteria governing PM approval or refusal of authorisation requests. An institutional head whose classification request is refused — or simply not decided within any defined period — has no apparent recourse under this text. The seven-day window applies to AKSIK’s administrative relay function, not to the PM’s substantive decision. What this produces is non-reviewable executive discretion over the boundary of secrecy itself. The PM determines, without defined criteria and without a required decision timeline, which institutions may classify and which may not. That gap predated the draft. The draft compounds it by adding a new layer of discretionary approval without supplying the criteria that would make that discretion accountable.

The Declassification Carve-Out and What It Leaves Unresolved

Article 22.6 of the parent law states a foundational principle: the classifying authority is simultaneously the declassifying authority. The institution that stamps information as secret is the institution that holds the authority to remove that designation. The draft inserts a new Article 22.6/1 that creates a direct exception to this principle.

Under the new provision, an institution that produces classified information may declassify it without being the original classifying authority, when doing so in response to classified information administered by another institution. The operational logic is identifiable. Inter-institutional classified exchanges can be obstructed when an institution needs to respond to a counterpart’s classified holdings without being able to reference or release its own classified sources. The amendment attempts to resolve a genuine friction point.

The parent law already delegates the rules governing declassification to Council of Ministers regulation under Article 22.7. That delegation is real, and it means the new exception does not land in a completely bare procedural landscape. What it does land in, however, is a gap in that delegation’s coverage. The draft’s implementation article — Article 11 — tasks the Council of Ministers with producing subsidiary regulation under Articles 2, 6, and 7 of the amending law. It does not task the Council with producing regulation specifically governing the new Article 22.6/1 exception. Whether the existing declassification regulation under Article 22.7 of the parent law will be interpreted to cover the new exception, or whether the exception will operate without specific procedural scaffolding, is left to administrative interpretation rather than legislative direction.

The provision is written in a single sentence. It contains no notification requirement toward the original classifying authority, no AKSIK oversight trigger, and no threshold for what constitutes a sufficient nexus between the information being declassified and the information being responded to. The structural consequence is that information ownership detaches from classification authority without a defined control loop. An institution may declassify material it produced, on the basis that it is responding to another institution’s holdings, with no mandatory record of that decision reaching either the original classifying authority or the oversight body. Whether this provision functions as a narrowly scoped operational tool or as a broadly available mechanism for managing institutionally inconvenient classified holdings will be determined not by the law, but by whoever chooses to use it. The legislative gap should be closed before the law takes effect.

AKSIK’s Enforcement Expansion and Its Structural Limits

The amendments spanning Articles 7 through 10 touch inspection principles, the administrative offence catalogue, and the enforcement architecture. Taken together, they represent the most significant expansion of AKSIK’s institutional position in the draft. The amendments to Article 70 codify inspection principles — independence, proportionality, programmatic planning — and align the procedural framework with the general inspections legislation, which the parent law had referenced but handled less precisely. Article 75 adds the physical transport of classified information to the administrative offence catalogue and inserts a new provision making AKSIK’s administrative violation decisions directly enforceable as executive titles.

The executive title provision is the sharpest instrument in the draft. Under the parent law, enforcement of AKSIK’s administrative decisions required non-compliance to be pursued through standard administrative channels. The draft changes this: an AKSIK decision on an administrative violation now compels compliance directly, with recourse available only through appeal by the sanctioned party. This is consistent with how a number of Albanian regulatory bodies operate. But the context here is classified information, where the underlying facts of any given violation may themselves be classified, and where the sanctioned party’s ability to mount a public defence is structurally constrained. That context matters when assessing what it means in practice to concentrate enforceable power over classified information in a single authority operating under conditions of limited transparency.

The most significant structural change in this section, however, is the reformulation of Article 78. Under the parent law, that article consisted of a single sentence: a person against whom an administrative measure has been imposed has the right to appeal to the competent court. There was no internal review mechanism. The draft replaces that article with a three-part structure. Violations are constatated by AKSIK’s inspection corps. Appeals against the inspection corps’ findings are heard by a separate structure established within AKSIK. Only after that internal process does the right of appeal to the competent court apply.

The result is that AKSIK becomes investigator, enforcer, and first-instance reviewer within the same institutional ecosystem. The creation of an internal review layer is not inherently objectionable — in a well-designed system it can provide faster, technically informed review before disputes reach the courts. But the draft establishes this structure without specifying how its independence from the inspection function is guaranteed. The inspection principles articulated in the amended Article 70 govern the inspection corps. They do not, by their terms, govern the new appeals body. Whether the head of the appeals structure is appointed independently of the inspection directorate, whether the same personnel may serve in both functions, and whether the appeals body has access to the full classified record of the underlying inspection are questions the draft leaves entirely to Council of Ministers regulation. For a body reviewing the decisions of its institutional sibling, on matters that may not be publicly discussable, these are not administrative details. They are the structural conditions of legitimacy.

The Emergency Destruction Protocol

Article 5, amending Article 38, is the most straightforward provision in the draft and the least controversial. The parent law requires public institutions to take measures for the evacuation of classified information during conditions of war, national emergency, or civil disaster. It does not address the scenario in which evacuation from a diplomatic mission abroad is impossible. The draft fills that gap, authorising mission heads to order emergency destruction of classified material when evacuation cannot be carried out.

The provision is appropriately scoped. It applies only in defined emergency conditions, requires a specific order from the mission head, and governs a situation where the alternative — permitting classified material to fall into hostile hands — is self-evidently worse than destruction. No analytical objection attaches to this amendment. It is competent legislative housekeeping that the parent law should have included from the outset.

The Implementation Architecture

Article 11 gives the Council of Ministers one year from the law’s entry into force to issue subsidiary regulation under Articles 2, 6, and 7 of the amending law. The one-year window is standard practice in Albanian legislative drafting. What is not standard, and not defensible, is the absence of a parallel obligation covering the new Article 22.6/1 declassification exception and the new internal appeals structure under Article 78. Both provisions generate accountability questions that the primary text does not resolve and that the existing regulatory framework under the parent law does not automatically cover.

The pattern across the draft is consistent. Provisions that expand executive discretion or create new institutional powers are either accompanied by delegated regulation mandates or contain their own self-executing procedural logic. The provision that carves out a new exception to the foundational principle of classifying-authority primacy, and the provision that inserts a novel internal review body into AKSIK’s enforcement architecture, are left without specific regulatory tasking. That asymmetry is not accidental drafting. It is a legislative choice, and it is the wrong one.

Conclusion

These amendments are not a security crisis. Several of them are defensible responses to real operational problems. The emergency destruction protocol is sound. The codification of inspection principles is overdue. The executive title provision for AKSIK decisions is consistent with the broader administrative enforcement framework, even if its application in the classified information context warrants careful subsidiary regulation.

The draft fails, however, to meet the standard of legislative completeness that its subject matter demands. The deepening of PM control over classification authority compounds a pre-existing accountability gap without supplying the criteria that would make that discretion reviewable. The declassification exception in the new Article 22.6/1 severs information ownership from classification authority without a defined control loop, in a single sentence, with no procedural mandate to supply one. The new internal appeals structure within AKSIK is created without the independence guarantees that would make it a genuine check rather than an institutional formality. Taken together, these omissions follow a common logic: the provisions that constrain institutional actors are procedurally detailed; the provisions that expand executive and regulatory discretion are not. That is not a drafting oversight. It is a structural signature.

Renada Bici is a Tirana-based lawyer practicing in civil, criminal, and administrative law. She holds a law degree from the University of Tirana and has experience in both private legal practice and public administration. She writes for the Tirana Examiner Legal Desk.