The content cut-off explained: which publications count for your competition

The rule

The Commission's guidance to candidates makes two promises about content. In paraphrase: every question rests on official EU sources that are publicly available, so that colleagues who left the Commission after applying are not at a disadvantage; and the test does not include publications released after the start of the application period.

The copy of the guidance we have seen is the AD7 version, and it names the date: 31 March 2026, the day COM/AD/107/2026 opened. The other 2026 competitions opened later, so applying the same rule moves their cut-off later too.

Why the rule exists

The Commission has not said. The record suggests an answer. After the March 2025 papers, a Commission staff union reported that at least one version of the AD6 paper carried a very high number of questions on the Draghi report, which it said is not an official EU policy document, and that one question on cutting administrative burden for SMEs was keyed to a figure the report recommended rather than the target the EU adopted. A rule limiting questions to official public sources, frozen at a known date, closes both gaps. That is our reading.

The cut-off for each competition

The estimates will hold until the AD5, AST3, AD6 and AST2 versions of the guidance are seen. If your guidance gives a different date, it wins. Each competition page carries its own cut-off: AD7, AD5, AST3, AD6, AST2 and EPSO/AD/427/26.

What falls between the cut-offs

Most of what was published between April and October 2026 changes nothing for a candidate. A few documents do. Each row below is checked against the official text. The omnibus packages are set out in simplification and the omnibus packages.

COM(2026) 380 is the one most likely to be tested, because it changes a number in the better-regulation rules:

Note the condition: the reduction applies when other consultations are run for the same initiative. For AD7 the 12-week standard stands alone; for the later competitions both the standard and the possible reduction are fair game.

The customs reform shows how one file can sit on both sides of a cut-off. The Commission proposed it in 2023, so the proposal is in scope for every 2026 competition. Its adoption is in scope only for AD6 and AST2:

Re-issued fact sheets: old facts, new dates

The European Parliament re-issued many of its fact sheets in spring 2026. Sheet 1 and sheet 7, for example, now carry the date 05-2026, and sheet 25 carries 04-2026. The facts inside are mostly long-standing, but the documents themselves are dated after AD7's cut-off. For a fact on composition, powers or procedure, the safer source is the Treaty article the fact sheet summarises: it is dated long before any cut-off, and it is what the examiners can rely on. That is how our questions are sourced.

Older documents still count

The rule excludes later publications, not earlier ones. A document from years ago that is still Commission policy stays in scope. Three examples that are easy to forget:

• the Energy Union Framework Strategy, COM(2015) 80, of 25 February 2015;
• the Zero Pollution Action Plan, COM(2021) 400, of 12 May 2021;
• the Commission's own digital strategy, published on 30 June 2022.

Pages that are updated continuously, such as the Commission's priorities page, are a grey zone: only what can be dated before your cut-off is safe. The high-yield reading list ranks what to read first, and the 2025 packages dates each package of the mandate.

EPSO works differently

For EPSO/AD/427/26 Part 2 there is no cut-off rule of this kind. EPSO says it will publish references to the source materials used to develop the test approximately two months before the testing session. None had been published by 26 October 2026. Once the list is out, it defines the corpus. See the EPSO AD5 Part 2 guide.

How we date our questions

Every question on this site cites one official source with its date, and carries a yes or no flag for each 2026 cut-off: AD7, AD5, AST3, and AD6 with AST2. The source register can be filtered the same way, so you can see what counts for your competition. When a source turns out to be dated or keyed wrongly, the change goes in the public corrections log. Try the free test to see the source and date under each answer.