As broadband internet access begins to take hold in Africa, this brings both opportunities and risks, which suggests paying attention to informational risks: cindynic methodology describes these risks as crucial in the construction of perceptions, and therefore in the genesis of conflicts, which hinder sustainable development, and even just development.

Second-order Cindynics enable a transdisciplinary and trans-sectoral approach of risks-conflicts-development complex, in particular thanks to the modeling of the concept of conflictuality, which is seen as a continuum. In this intangible and organisational approach of situations dynamics, informational risks play an essential role, and are divided into three classes: opening risk, closing risk and toxic risk.

The concept of opening describes access to, or even disclosure of, confidential information. The notion of closing corresponds to the blocking of the publication of public information, or attempts to blacken white information. And the notion of toxicity characterises the dissemination of information during deception operations targeting either individuals or entire populations, more or less accidental disinformation, or even the dissemination of messages generating vulnerabilities or increasing conflictualities.

Example of opening risks: the disclosure of the real names of Facebook users, imposed by its real name policy. This policy endangers lives, as the recent US debacle in Afghanistan demonstrated.

Examples of closing risks : Censorship imposed by Big Techs like Facebook or Twitter, which threatens all users in general, journalists in particular, and even Presidents of African countries. The Nigerian President was thus censored last June. Another example: the obligation to take down content imposed on French hosting providers since the transposition of Directive 2000/31 EC, which requires them to censor content before any trial. This obligation of censorship by service providers was then conceptualised as "self-regulation". If the European Union promotes this self-regulation, this could facilitate the dissemination of the legalisation of this practice, including in Africa.

Examples of toxic risk: In Rwanda, in 1994, calls for the elimination of Tutsis by radio 1000 Collines. Or around the same time, CNN editorials opposing a UN intervention in Rwanda. More recently: deception operations carried out in Africa by the trolls of Yevgeny Prigozhin conglomerate. Or the statements of Vincent Duclert following the publication of the report on Rwanda ordered by the French presidency, which de facto extol the actions of armed groups.

These risks are interrelated: for example the censorship by Big Tech of opinions expressed on social media is a closing risk which brings toxic risks, since censorship can lead to a distortion of reality and manipulation of opinions, possibly intended, in particular during election periods. This brings a risk of election flipping, or even civil war.

The opening risk generated by Facebook's real name policy does not only endanger lives, it also generates a risk of censorship, through the phenomenon of chilling effect : users who are aware of the risks associated with the inability to be protected by pseudonymity will self-censor. This opening risk therefore brings a closing risk. Which in turn brings a risk of distorting reality, i.e. a toxic risk. 

The toxic risk linked to the activities of the Prigozhin conglomerate brings a closing risk: In order to fight against these activities, Facebook censors content, invoking a concept of "coordinated inauthentic behavior" based on its real name policy. This generates a closing risk since in the name of this concept, Facebook censors actors who fight against disinformation. In addition, this toxic risk itself generates another toxic risk: by fighting against  a Russian deception operation, Facebook implied that the French army was behind the accounts fighting against these Russian operations. Hypothesis then taken up as a fact by the media: Facebook therefore generates disinformation.

In addition, deception operations on the Internet, be they carried out by state or non-state actors, encourage states to adopt legislative measures extending Internet surveillance. In the case of France, all of the connection activities of French Internet users are systematically logged a priori. This constitutes a considerable opening risk since it is the entire population that is placed under surveillance: in comparison, the tens of thousands of people eavesdropped in a targeted manner by the Israeli software Pegasus appear to be totally negligible, although Ursula von der Leyden vehemently criticised the use of Pegasus. Moreover, this illustrates the cindynic notion of cindynic degeneracy: the media fulminate against the eavesdropping of a few tens of thousands of people punctually monitored, and do not mention the tens of millions of French people whose internet connections are systematically logged ex ante

Basically, all of these issues point to three cindynic notions. First of all, the notion of an invasive limit: prevention involves observing situations. But how far can prevention go in acquiring information on the actors in a situation? At one point, this acquisition becomes invasive, and itself causes damage by violating confidentiality. So there is a fundamental limitation to prevention, a kind of uncertainty principle, according to which perfect prevention would cause damage, and the protection of confidentiality would restrict prevention. The question is: where is that invasive limit, and who can say that.

Which leads directly to another cindynic notion: the legitimacy diagonal. Second-order Cindynics consider spectra of situations differently perceived by different observers, who have divergences on the transformations that should be applied to a situation. This means in particular that each actor wants specific transformations to be applied to other actors. The notion of diagonal of legitimacy illustrates that in a situation spectrum, each actor has the legitimacy to transform himself, but does not have the legitimacy to transform others as he wishes.

Finally, this last notion refers to the notion of diversity: in the cindynic sense, diversity is the state of a situation where actors are different, but where they have no divergences, and do not wish to change other actors in a way that is not the way they want.