Quad9, the Swiss non-profit behind the free 9.9.9.9 DNS resolver, will stop serving users in France on Tuesday, 24 November 2026. The trigger was a French court ruling on 21 October. It left the organisation choosing between rewriting DNS answers for its worldwide user base and defying an order it says it cannot comply with only inside France. Quad9's post, "Farewell, France", says roughly 20 further lawsuits are pending against it and other resolver operators.
The strongest case for the orders
Rights holders have a real grievance. Live sports is the one content category where piracy destroys value within hours, and broadcasters pay heavily for exclusivity. Blocking by ISPs is easy to evade. A user can switch to a public resolver in about a minute. From a rights holder's view, a resolver that ignores a court order is a loophole that defeats the ISP blocks.
France has built a legal framework around that argument. Article L.333-10 of the Code du sport lets a court order "any person likely to contribute" to ending an infringement to take proportionate measures, including blocking, for each matchday over up to twelve months. Légifrance shows the article as amended by the law of 3 August 2026 and in force since 5 August. It also lets the media regulator Arcom pass identified sites to intermediaries, with fines of up to 1% of worldwide turnover for non-compliance. I have not verified which provision the 21 October judgment rests on, but this is the framework French sports-piracy orders are built on.
What actually happened to Quad9
Per Quad9's account, the broadcaster beIN sued over audiovisual rights in sports broadcasts. The court "significantly reduced the financial penalty, from €580,000 to €1,000 per day." Quad9 puts the exposure at €24,000 if it does not comply until the blocking ends on 14 November. TorrentFreak reports the order enforces an earlier May blocking order that has since grown to 58 domain names.
The penalty is small. The pending cases are the threat. Quad9 says it "cannot survive the accumulation of penalties yet to be assessed in 20 other pending cases," which together demand it block hundreds of names. An appeal against the 21 October ruling is pending, but Quad9 is leaving France in the meantime.
Why a resolver cannot localise a French order
The order's premise is that blocking can be confined to France. A free, no-registration resolver has no reliable way to know where a query comes from. Quad9 says it has no geographic blocking methods, and that geo-IP blocking is easily bypassed by VPNs. Its other option is to add the domains to its global malicious-domain list. It calls that "entirely unworkable" because French law would then apply to users everywhere.
This is not new. In December 2024, Quad9 said it faced a similar Canal+ order and noted that Cisco had "exited the French market for DNS entirely" after earlier demands. Each exit leaves French users with fewer privacy-respecting resolver choices, and the usual fallbacks are their ISP's resolver or a large US platform.
The enforcement logic does not hold
DNS blocking does not remove any content. The pirate sites keep running, and anyone with an IP address, a VPN, or a different resolver reaches them. As the Electronic Frontier Foundation notes, the withdrawal "will not remove any allegedly infringing material from the internet." The measure's costs fall on the compliant, lawful intermediary and its users. Its benefits are partial at best.
There is also a structural problem. A resolver is a general-purpose infrastructure layer. Once courts treat it as a blocking point, every rights holder, in every sector and country, has the same argument. Quad9's warning that "once the tools of censorship are built, they are rarely used sparingly or justly" is a prediction about incentives. It is not a claim about this particular case.
Proportionality also gets short shrift. L.333-10 itself requires "proportionate" measures. Ordering a non-profit with no French subsidiary to apply French law to a global service, with penalties that multiply across dozens of suits, is hard to square with that standard. The measure is only proportionate if the target can comply narrowly, and here it cannot.
A better design
France already has a mechanism that works better. Arcom's identification system lets rights holders report live-event pirate sites to the intermediaries with direct control over delivery: ISPs, hosts, and content delivery networks. Enforcement should concentrate there. Where a resolver can credibly localise its response, as large operators with geographic infrastructure may, courts could include it. Where it cannot, as with a small global non-profit, the order should not demand impossible geographic precision, and penalty accumulation across parallel cases should be coordinated or capped so that the aggregate stays proportionate.
The case for protecting live sports rights is legitimate and strong. The remedy still needs to fit the infrastructure. Pushing out a privacy-preserving, non-commercial resolver weakens the open internet in France and does little to the pirate streams it was meant to stop.