Summary · English

Neither shadow government nor charitable club

This investigation takes seriously a question the Belgian debate handles badly: does freemasonry shape public decisions? It refuses both easy answers — conspiracy and denial — and grades every claim by what the sources actually establish.

The verdict in one sentence

The political influence of Belgian freemasonry is historically demonstrated for the nineteenth century, particularly in the construction of anticlerical liberalism and secular institutions; today the public record establishes a pluralist milieu of socialisation, networking and occasional advocacy — but no single hidden centre directing Belgian politics.

Six obediences, not one bloc

The fault line does not run between masons and non-masons. It runs between two conceptions of what a lodge is.

  • The regular current requires belief in a Supreme Being, admits men only, and statutorily forbids political, religious and societal discussion in lodge. In Belgium it is carried by the Regular Grand Lodge of Belgium (≈ 1,900 members), the only Belgian grand lodge recognised by the United Grand Lodge of England. That recognition is an inter-masonic label — not a public accreditation, not a quality rating, and of no legal consequence whatsoever.
  • The liberal or adogmatic current is overwhelmingly the majority: Grand Orient of Belgium (≈ 10,000), Belgian Federation of Le Droit Humain (≈ 7,000, mixed, about 74 % women), Grand Lodge of Belgium (≈ 3,800), Women's Grand Lodge of Belgium (≈ 2,250) and Lithos (≈ 1,100). It is not a doctrinal or political bloc: the disagreements are real.

In July 2025 several obediences published a joint statement against a draft bill authorising home visits to arrest people without residence permits. The Regular Grand Lodge did not join — consistently with its doctrine — and said so publicly. A unanimous position of "Belgian freemasonry" has never existed and cannot exist.

What the lodges did, and what they did not

They did: found, in 1834, the institution that became the Université libre de Bruxelles, with lasting funding from masonic circles; provide the relational backbone of part of Belgian liberalism after the episcopal condemnation of 1837; serve as the meeting ground between liberal and socialist elites that made the 1958-59 School Pact possible; and incubate the Flemish Humanistisch Verbond in 1951 with premises and money before it became fully autonomous.

They did not: organise the Belgian revolution of 1830. The research of Anaïs Maes and Jeffrey Tyssens finds no coordinated institutional action whatsoever; lodges and members were split between Belgicists and Orangists. Individual freemasons took part in the events — that is not orchestration. This is the most widespread claim of the popular narrative, and the easiest to falsify.

Seven mechanisms, one real problem

Ideological socialisation, networking, expertise and public advocacy are legitimate — comparable in principle to a church, a trade union, a university or an NGO. Targeted, non-visible contact with elected representatives — the January 2018 letter from the Khaos lodge to "brother" parliamentarians — sits at the boundary. Patronage in appointments or public procurement, and coordinated voting instructions, are not acceptable — and there is no evidence of either in Belgium.

The asymmetry matters: the best-documented mechanisms are the least dangerous, and the most dangerous are the least documented. That may mean they are rare. It may also mean that nothing in Belgium is designed to detect them.

What the law says

The Belgian federal code of conduct defines a conflict of interest functionally: any personal interest liable to influence the impartial exercise of a public function. Membership as such — philosophical, religious, trade-union or masonic — is therefore not enough.

The European Court of Human Rights has held the same since Salaman v. United Kingdom (2000), where a complaint based on a judge's shared masonic membership, unconnected to the subject matter of the dispute, was declared manifestly ill-founded. On 7 July 2026 the Grand Chamber held in Grande Oriente d'Italia v. Italy that the seizure of membership lists containing the personal data of more than 6,000 individuals, without adequate safeguards against arbitrariness, violated Article 8. Note the scope carefully: the case concerns a search and a mass seizure at an association's premises — not a blanket prohibition on any targeted declaration duty.

A public register of members would therefore be disproportionate. But a blind spot remains: if a judge must rule on a member of his own lodge and does not recuse himself, a litigant has no lawful means of proving the connection. The problem is not one of proof but of the appearance of impartiality protected by Article 6 § 1.

The real Belgian failure

GRECO, the Council of Europe's anti-corruption body, found in its fifth-round addendum published in December 2025 that Belgium had satisfactorily implemented only 8 of 22 recommendations (10 partially, 4 not at all). The gaps concern lobbying contacts, ministerial cabinet staff, asset and interest declarations and their verification, gifts, and post-mandate employment.

This concerns every organised interest, not freemasonry in particular. But it is exactly why nothing can be demonstrated — in either direction. Regulate the activity, not the identity: a mandatory legislative footprint, published agendas of substantial contacts, verified interest declarations covering leadership positions, and documented recusal.

Read on

The full site is in French: the six-chapter investigation, the register of 21 claims graded by evidence level, the critical audit of the sources, and the open data (CC BY 4.0). A Dutch summary is also available.

Written, fact-checked and coded by Claude (Anthropic), August 2026. The commissioning party has chosen to remain anonymous.