According to the annual report of the DGCCRF (Directorate general for Competition, Consumption and Repression of fraud) published Monday, July 27, more than 55,000 complaints from consumers have been identified regarding subscriptions for subscription hidden, abuse of weakness in the troubleshooting at home, fake administrative sites, and especially in the direct selling abusive.
To respond to these scams, in addition to more controls by targeting donors, triple the amount of the fines and expand the “name and shame” (systematic for an offence of the type the delay of payment ), the DGCCRF now uses artificial intelligence to check the fake reviews.
The complainers at the origin of most of the false reviews
Why fake online reviews are they so widespread? Since the widespread use of e-commerce, dissatisfied clients have used the opportunity to share their comments to express their dissatisfaction to the four corners of the world. Negative opinions can spread more easily than positive reviews and the customers are more likely to express themselves in case of a bad experience, which can have very negative consequences for a brand or a service, especially if the opinion was exaggerated, unwarranted, or if the company has since corrected its errors.
Over 53% of individuals recognize to be influenced rather by the opinion of the consumers by search engines and promotional e-mails.
The difficulty to manage and generate positive feedback drives traders and even service agencies (travel, real estate…) to buy favorable reviews. According to Virginia Beaumenier, director of the DGCCRF some also resort to “agencies” of online reviews to create positive opinions, but also to denigrate their competitors.
This practice is punished by the law, by penalties ranging up to 300,000 euros and two years in prison.
The investigators will be assisted by algorithms
To identify the scams for sure, the DGCCRF does not rest solely on new technologies. The repression of the frauds first sends its investigators on the ground, to collect data, and verify the actual quality of the services offered. It is in a second time involved machines: algorithms automatically identify the fake notices, which do not correspond to the reality observed by the investigators, and which abound on the sites of online trading and on platforms, hotels, or restaurants.
The DGCCRF has identified some criteria to help identify fake reviews in a faster way. Accounts of the origin of a large number of comments were also examined in an automatic way to detect the opinion purchased by traders.
The automation of contracts, estimates and invoices
The artificial intelligence will also be applied to the systematic analysis of the contracts, quotes and invoices. Algorithms will help identify the clauses and abusive practices to enable officials to better target their controls.
This is not a first in terms of artificial intelligence to enforce compliance with the act. The Ministry of public finance has also put in place a technology of “data mining” to analyze colossal amounts of information on businesses and individuals, and identify sales hidden, the real estate under-valued or cash flow on accounts of the suspects.
The departments of the Drôme, of the Charente-Maritime and the Maritime Alps have also used artificial intelligence to review satellite images and more easily locate the pools of undeclared or illegally built.