Since 2020, an experimental law allows the tax authorities to consult social networks and major web platforms to detect fraud or violations of tax obligations. The association La Quadrature du Net challenged this law, which it considers too intrusive, during an appeal rejected by the Council of State.
Social networks are a goldmine of information about our consumption habits. You may not know it, but since 2020 an experimental law allows the tax authorities to consult social networks and certain platforms for the purpose of detecting fraud.
Article 154 of the Financial Act explains that on an experimental basis “the tax authorities and the customs and indirect rights administration may, each for their own part, collect and exploit the content by means of automated and automated processing (…) the content, freely accessible on the websites of the platform operators in line that are clearly public made by their users”. The aim of this experiment, which is spread over three years, is to detect fraud such as illegal sale of products, false domiciliation abroad or tax evasion.
La Quadrature du Net, an association for the defense of freedoms on the internet, recently filed an appeal with the Council of State to have its application annulled. The association denounces the implementation of generalized and undifferentiated surveillance, authorizing the two administrations to collect data from citizens on a massive scale.
The nature of the data collected at the heart of the profession
“Such a device directly and seriously affects the exercise of fundamental rights in the digital environment and individual freedoms in the field of automated data processing, exposing the population to unlawful surveillance”, noted La Quadrature du Net during her profession. The public institution sided with the executive and did not oppose the law.
However, the association La Quadrature du Net is concerned about issues already raised by the CNIL during the first agreement on the legislation. Several concerns had been raised, in particular about the contours of the device, such as the “freely” accessible nature of the targeted content, the nature of the data that could be collected or even the guarantee that only the data needed can be used to detect fraud.
Since 2013, the administration has been developing automatic data processing software to improve fraud detection and facilitate the targeting of tax audits. The “targeting of fraud and valuation of requests” (CVFR) is based on data mining techniques, by crossing data from different sources of the tax authorities and open access data. For its part, the experiment should soon end and no one knows yet what the future of this law is.
ETX Daily Up