Can We Trust Social Media Data? Social Network Manipulation by an IoT Botnet.

Social media fraud is generating likes, friends, views or any other action on social media services to artificially increase a following. While it offers social media users a cheap and easy means to approximate digital influence, it generates considerable costs for social media providers and funds botnet operators. In addition, it erodes the trust of its users.

Paquet-Clouston, Bilodeau and Décary-Hétu sought to gain a deeper understanding of the illicit market for social media fraud. They evaluated the supply of social media fraud services by analyzing their availability and cost. They deployed machines inside the Linux/Moose botnet to observe its operations and profile its customers. It is one of the first-time researchers have analyzed social media fraud from the inside of a botnet.

The team queried Google between January and December 2016 and found over 5,600 ads for social media fraud services. During that time frame, the team collected traffic internal to the Linux/Moose botnet, a botnet that connects to social media services through compromised routers and Internet of Things (IoT) devices. They collected over 273,000 transactions. Most of these transactions were directed at Instagram. By analyzing the transactions, the researchers identified 522 accounts that seemed to have used social media fraud.

The study revealed that social media fraud services are varied and readily available. They are offered at a variety of prices which makes them accessible to any type of consumer. Prices vary depending on the targeted social media platforms. It appears that services that are more difficult to defraud are more expensive. Instagram services were generally less expensive. The researchers also found that the customer base for such services is diverse. Individuals make up the largest category of consumers, followed by companies and entrepreneurs. The fake-user account detection measures on Instagram generally detect and shut down fake accounts within six months of their creation. However, these measures do not prevent Linux/Moose from operating.

The researchers suggested that tackling this issue would require actions on several fronts, including:

• Designing better registration and account-creation barriers.
• Provide social media users with legitimate and affordable solutions to grow their fan base.
• Raise awareness among social media users of the illicit nature of social media fraud services.

Reducing social media fraud requires action from a wide range of actors including consumers, social media platform owners, law enforcement agencies and service providers.

Vous pourriez également aimer

Mots-clés