Following the looting over the weekend, the role of social media in coordinating rioters is at the heart of the debate. Gabriel Attal has called for certain platforms to be blocked temporarily and in a targeted manner. This proposal revives a sensitive question: how far can authorities restrict digital services during periods of violence without disproportionately infringing on freedom of expression and access to information?
Riots: could social media be restricted or shut down?
The proposal raised after the looting aims to limit access, for a set period, to certain platforms used to share information or organize violent gatherings. Presented on BFM Économie’s Good Morning Business show, it comes at a time when authorities are seeking to prevent calls to gather and the rapid spread of instructions related to the riots.
This does not necessarily mean shutting down the entire internet. The debate is instead about targeted measures: slowing down, suspending or temporarily making certain digital services unavailable, or even limiting specific features. This distinction is important, because a general network shutdown would also affect residents, emergency services, businesses and media outlets with no connection to the violence.
Why are social media platforms being blamed?
During periods of unrest, platforms make it possible to post videos, addresses, times or calls to join a location in just a few seconds. This content can be shared quickly and reach users who do not know one another. Authorities fear that this accelerated circulation could facilitate the formation of groups, the transmission of instructions or the spread of rumors that could worsen tensions.
However, rapid dissemination is a general feature of social media, not proof that every post contributes to the organization of illegal acts. People may also use these platforms to seek information about the situation, report a danger or document events. An overly broad restriction could therefore deprive the public of useful communication tools at the very moment when reliable information is particularly necessary.
Different uses within the same platform
A single app can host calls for violence, eyewitness accounts, messages from loved ones and official information. Problematic content is not always easy to distinguish in an emergency, especially when it is posted as images, videos or coded messages. A complete service block can therefore have much broader effects than removing posts directly related to the violence.
Platforms can remove content, suspend accounts or limit the visibility of certain posts. But these interventions depend on their moderation tools, internal rules and ability to process a high volume of messages quickly. Authorities, for their part, can request cooperation or consider technical measures, but the existence of a political request does not automatically mean that a shutdown will be implemented.
Temporary and targeted blocking: what could it mean?
The terms “temporary” and “targeted” encompass several possibilities. The measure could target a particular platform, a broadcasting feature or certain content, rather than all services accessible from within the territory. It could also be limited to a specific area or period, for example when a risk of further violent gatherings has been identified.
Each option raises practical challenges. A geographic restriction requires precisely determining the areas concerned and avoiding affecting nearby users without cause. Disabling a feature may be more limited, but depends on how the platform is designed. Finally, a general shutdown of a service may be circumvented by some users while penalizing those who use it for everyday activities.
Technology does not always allow for a perfectly selective shutdown
Major platforms rely on multiple infrastructures and functions: messaging, video posting, live streaming, alerts and private exchanges. Blocking a tool precisely can therefore prove more complex than a simple administrative decision. Depending on the methods chosen, access may be slowed, partially interrupted or completely unavailable, with varying effects depending on devices and service providers.
Users may also try to circumvent restrictions by using technical tools or alternative services. This does not mean that a measure would necessarily be ineffective, but its actual impact would depend on its duration, scope and the authorities’ ability to prevent circumvention. A restriction that significantly disrupts communications without reducing violence could fuel controversy rather than address the emergency.
Freedom of expression and oversight of measures
Restricting access to social media affects several freedoms, including freedom of expression, the right to receive information and the freedom to communicate. In a state governed by the rule of law, a restrictive measure must be subject to safeguards and justified by a specific objective. It must also be proportionate to the risk, limited in time and open to review or challenge.
The central question, therefore, is how to balance the prevention of violence with the protection of fundamental rights. A general measure could affect millions of people who are not participating in the riots. Conversely, action targeting identified content may be more proportionate, but requires the ability to quickly identify the posts in question and intervene without obstructing legitimate exchanges.
Online incidents also show that identifying an account or a post is not always enough to establish the account holder’s responsibility. An article about a Norman man facing prosecution for condoning terrorism, for example, discusses the possibility that his account had been hacked. Such cases highlight the importance of checking the technical evidence and context before drawing conclusions about who actually authored a message: the case of the account that may have been hacked.
Foreign precedents fuel the debate
The issue of restrictions is not limited to periods of riots. In several countries, governments are debating age limits, the protection of minors and platform liability. These discussions have different objectives, but they reflect a broader trend: authorities are seeking to impose greater regulation on access to social media and the effects of its use.
In Australia, the ban on social media for teenagers was presented as an experiment to be assessed. The issue highlights the difficulties involved in age verification, effective enforcement of the rules and the consequences for young people who use these tools to communicate or get information. This measure is reviewed in this article on Australia’s gamble on teenagers’ access to platforms.
New Zealand has also considered restricting access for those under 16, in the name of child protection. This proposal shows that restrictions can target specific groups or uses rather than all users. It nevertheless raises practical questions: how can age be verified, which platforms would be affected and how can young people be prevented from moving to less regulated services? These issues are explored in this article on New Zealand’s proposal targeting under-16s.
In France, the Senate is also set to debate a possible ban on social media for those under 15. This issue primarily concerns the protection of minors, but it raises questions similar to those in the debate on riots: the scope of the restriction, its legal basis, its technical implementation and the safeguards required. The parliamentary timetable is outlined in this article on the Senate debate concerning under-15s.
Moderation, censorship and platform decisions
The ability to quickly remove content or an advertisement also depends on the choices made by platforms and local authorities. These decisions are not limited to periods of crisis: they concern the visibility of posts, commercial rules and compliance with national laws. An article about a Solana advertisement that was removed after just a few minutes in China, and the criticism directed at CZ, illustrates these tensions surrounding content control and how decisions are attributed: the example of the Solana advertisement that was quickly pulled in China.
During riots, authorities can ask digital companies to remove posts or provide information in accordance with the applicable procedures. Platforms must then balance their legal obligations, moderation rules and the urgency of the situation. An intervention that is too slow can allow dangerous calls to circulate; one that is too broad can remove content of public interest or prevent eyewitness accounts from being shared.
Clear criteria to prevent excessive restrictions
For a measure to be understandable and subject to review, its criteria should be clearly defined: which services are affected, what events justify it, how long will it last and what conditions will bring it to an end? Transparency about these points makes it easier to assess its effectiveness and distinguish a temporary response to an immediate risk from a broader restriction of the digital sphere.
Accountability is also essential. A platform cannot always anticipate the spread of content, particularly when it is posted in a private group or in a form that is difficult to detect. Authorities, for their part, must establish a link between the messages in question and the unrest observed. Quickly assigning responsibility to a service or a user can lead to contested decisions if the available evidence is incomplete.
What impact would this have on residents and essential services?
A shutdown or restriction can affect far more than people suspected of participating in the violence. Merchants may lose a channel for communicating with their customers, families may have difficulty finding one another and residents may lose access to local information. Associations, journalists and public services may also use the same platforms to share instructions or report emergencies.
Social media is not the only way to share safety information. Authorities can rely on official websites, broadcast media, phone alerts and local communication channels. But these alternatives are not always accessed in the same way, and their effectiveness depends on their availability, visibility and the public’s trust. Any restriction should therefore be accompanied by accessible alternative channels.
The debate is not simply a choice between taking no action at all and imposing a complete shutdown. Other responses could be considered: requesting the rapid removal of content calling for violence, strengthening cooperation with platforms, sharing verified information or prosecuting offenders through established procedures. Each of these options has its limitations, but combining them could prevent the entire response from resting on making a digital service unavailable.
An emergency measure that would remain controversial
Gabriel Attal’s call for a temporary and targeted block presents the authorities with a practical challenge: acting quickly in response to calls that could contribute to violence while avoiding a disproportionate restriction of communications. The possibility of intervening on certain platforms is a subject of political debate, but implementation would depend on the legal framework, the technical approach adopted and the safeguards put in place.
Further discussions will need to clarify, in particular, what the term “targeted” covers: an entire app, a specific feature, identified content or a geographic area. They will also need to establish how the measure’s effects will be assessed, how users will be informed and what avenues of appeal will be available. Without these clarifications, the debate risks conflating action against calls for violence with a general restriction on access to social media.
