In France, an algorithm runs to the aid of the treasury in the hunt for tax evaders. In fact, in a few weeks, an artificial intelligence will collect information from social networks and the web to find the crafty ones from the French tax authorities. Public Accounts Minister Grald Darmanin announced it via Twitter just over a year ago and now a decree published on February 13, 2021 specifies the terms of application of article 154 of the budget law for 2020, which gives rise to a widespread monitoring of social networks by the French tax administration. So, soon Twitter, Facebook, Instagram and all the other social networks used by the French will be scanned by an algorithm that will collect the information necessary to unmask tax evaders.
Three years of experimentation
The trial is expected to last three years, starting as early as March, according to reports Le Figaro, with tax agents who will be able to check social networks, but also websites such as Airbnb, and cross photos, keywords, dates and places thanks to an automatic learning algorithm and thus detect the presence of fraudulent activity. The French authority has already replied to those who promptly raised doubts about privacy, giving its ok to the operation, but on one condition: the data that can be collected are those published spontaneously by users and freely accessible on the platforms.
The Italian situation
And in Italy, where tax evasion is one of the worst scourges in the country, how does it work? In fact, we too can scan the web for tax evaders. As Alice Scaglioni wrote in the Courier service a year ago, a circular (no. 16 / E of 28 April 2016) of the Revenue Agency confirms it, which states that the information obtained from the databases can be added “those that come from other sources, there open “which could therefore include public data on social networks. Another document, circular no. 1/2018 (published on 4 December 2017) of the Guardia di Finanza, which speaks precisely of the possibility of looking for useful elements not resulting from databases, with particular attention to the consultation of “open sources” (press articles, websites, social network) in order to acquire every useful element of knowledge on the taxpayer to be controlled and on the activity exercised by them.