What the Federal Council decided
On 14 October 2026 the Federal Council took note of the consultation results on the proposed Federal Act on Communication Platforms and Search Engines (KomPG) and told the Federal Department of the Environment, Transport, Energy and Communications (DETEC/UVEK) to draft the dispatch to Parliament. Two changes matter. The preliminary draft is to be amended with child-protection measures, and the reporting procedure is to be extended to include all content suspected of being unlawful (Federal Council release, English).
The dispatch is due by the end of June 2027, according to the official admin.ch release and blue News. Blue News also reports that the government had expected the rules to take effect in 2029 at the earliest.
The consultation opened on 29 October 2025 and closed on 16 February 2026 (law.ch). According to blue News, 214 responses came in. Of these, 53 asked for all illegal content to be reportable and 12 wanted to keep the original list of offences.
The strongest case for the government's approach
The case for widening reporting is simple. Users do not know which of nine criminal-law categories a harmful post falls under. A narrow list invites platforms to reject legitimate reports on technicalities. The consultation shows demand, since respondents backing the broader mechanism outnumbered those defending the narrow one more than four to one. Child protection is equally easy to justify. Very large services must assess where they create or amplify risks for minors and mitigate them, and they may no longer show minors profiling-based advertising (blue News). Blue News reports that a ban on advertising to minors was the most requested child-protection measure, with 56 mentions. That ban targets a commercial practice, not speech, and it is the least contentious part of the package.
Where proportionality is at risk
A reporting channel is not itself censorship. The risk comes from what platforms do with the reports. The draft's other duties include telling affected users when content is removed or accounts are blocked, giving reasons, offering an internal complaints procedure, and taking part in out-of-court dispute settlement (law.ch). Those safeguards are what make the model defensible. They put process around moderation without requiring any particular outcome.
The original design tied the channel to specific offences, including defamation (Art. 174 StGB), insult (Art. 177 StGB) and incitement to hatred (Art. 261bis StGB). Courts have developed these provisions over many years. Insult and defamation in particular turn on context, intent and public interest. A user who reports a post as 'presumably unlawful' is asking the platform to make a legal judgement that Swiss courts often find difficult. Under a universal mechanism, a large share of reports will concern content that is offensive but lawful, or that is a lawful political attack on a public figure.
If each report triggers a duty to examine and respond, the cheapest compliance path is to remove when in doubt. The statute should say that a report does not oblige removal, that platforms may decline manifestly unfounded reports, and that repeat abusive reporters can be deprioritised. The dispatch should also publish how many reports led to removal and how many of those were reversed on appeal. Without these provisions, the broadened channel could turn a transparency law into a takedown pipeline.
The best feature remains its narrowness
The act applies only to services used monthly by at least 10% of Switzerland's resident population, currently about 900,000 people (law.ch; blue News). AI-powered search engines are now explicitly included, and smaller platforms stay out even though 14 cantons wanted them covered. That is the right call. Swiss startups and regional forums would otherwise face compliance duties built for global platforms. Switzerland has a small domestic market and a thin pool of in-house legal talent. The relevant comparison is the EU Digital Services Act, which the draft follows closely but does not copy. The Swiss law keeps its obligations proportionate by reaching only the largest services.
The AI-search inclusion deserves care. Rules written for ranked link lists do not map cleanly onto generated answers. Legislators should define precisely which systemic-risk duties apply to a generative search product and which are simply carried over from the platform model.
Child protection without a surveillance mandate
The open question is age verification. Blue News reports that the Federal Council will decide on further youth-protection measures after a report on postulates 24.4480 (Vara) and 24.4592 (Graf), with DETEC working alongside the Federal Department of Home Affairs and the Federal Social Insurance Office (BAKOM release, German). Mandatory age checks would require every adult to prove their identity to use mainstream services. That raises privacy and anonymity costs that risk-assessment duties and the advertising ban do not. The evidence from other jurisdictions is that verification regimes push users to unregulated alternatives and VPNs. Parliament should require the report to weigh those effects explicitly before any verification mandate is drafted.
What to watch
Three things will show whether the final text is proportionate. First, whether the dispatch limits the legal consequences of a report to review and a reasoned answer, not mandatory removal. Second, whether the transparency and researcher-access duties come with strong data-protection safeguards. Third, whether the age-verification decision is made on evidence rather than political momentum. Switzerland has the advantage of time, with the earliest entry into force around 2029. It should use that time to publish the removal and reversal data that would let Parliament test the broadened channel.