Author: Aaron Wood, CoinTelegraph; compiled by: Deng Tong,

Meta, parent company of Facebook, Instagram, WhatsApp and Messenger, is currently facing antitrust lawsuits that could limit its ability to develop artificial intelligence among many competitors.

The Federal Trade Commission (FTC) first filed a lawsuit in 2021, accusing Meta of its acquisition strategy, rather than competing with it, of violating antitrust laws. If the court ruled that Meta lost, it could be forced to spin off its various instant messaging services and social media sites into independent companies.

Losing its stable social media companies could not only undermine Facebook’s competitiveness in the social media industry, but also its ability to use the data from these sites to train and develop its proprietary Llama AI models.

The trial may take months to a year, but the outcome will have a lasting impact on Meta's position in the AI ​​competition.

Meta's antitrust case and its impact on artificial intelligence

The Federal Trade Commission (FTC) first filed a lawsuit against Meta in 2020, when the company was still operating under Facebook’s name. A year later, the agency's revised complaint alleged that Meta (then called Facebook) used an illegal "acquisition or burial" strategy against more creative competitors after "failed to develop innovative mobile features for its web." This has led to a monopoly in the social media market of “friends and family”.

On April 14, the first day of the formal trial of the FTC v. Meta, Meta founder and CEO Mark Zuckerberg had the opportunity to respond to the allegations. He testified that only 20% of user content on Facebook and about 10% of user content on Instagram are generated by users’ friends. Zuckerberg claims that the nature of social media has changed.

“People are just constantly involved in more and more things their friends won’t do,” he said — meaning Meta’s social media assets are diverse enough in nature.

The FTC accused Meta of discovering potential threatening competitors and buying them. Source: FTC

When the FTC first filed a complaint, Meta called the allegations “a history of revisionism” and reiterated the claim on April 13, saying the agency “ignored reality.” The company argued that the acquisition of Instagram and WhatsApp benefited users, and that YouTube and TikTok brought competition.

If the District of Columbia Circuit Court decides adversely on Meta, the global social media giant will be forced to split the services into independent companies. Jasmine Enberg, vice president and chief analyst at eMarketer, told the Los Angeles Times that such a ruling could cost Meta a competitive advantage in the social media market.

"Instagram is indeed its biggest growth driver, as it has long been making up for Facebook's shortcomings, especially with younger users," Enberg said. "Facebook hasn't been a place where cool college students gather in a long time."

Such a ruling will also affect the database Meta uses to train its AI models. In July 2024, Meta stopped launching artificial intelligence models in the EU on the grounds of “regulatory uncertainty.”

Meta paused the behavior after privacy advocacy group None of Your Business filed a complaint in 11 European countries about Meta using its platform’s public data to train its AI models. The Irish Data Protection Commission then ordered the practice to be suspended until review was conducted.

On April 14, Meta was approved to use public data (i.e. posts and comments from adult users on all its platforms) to train models. If these companies dissolve into independent companies with their respective organizational structures and data protection policies and practices, Meta will lose the data ocean and interpersonal communication that can improve its artificial intelligence.

In this case, Andrew Rossow, cyberspace attorney and CEO of AR Media Consulting, Minc Law, noted that in this case, “companies will likely control their own user data, and Meta will be restricted from using this data unless a new data sharing agreement is reached, which will be subject to regulatory scrutiny and user/consumer privacy laws.”

However, Rossow pointed out that this will not be a complete loss for Meta. Zuckerberg's company will retain a large amount of data from Facebook and Messenger. It can continue to use “opt-in” data that allows its posts to be used for AI training, as well as synthetic datasets as well as third-party and open data.

Meta, AI competitions and data protection

With DeepSeek joining the competition and Meta launching its fourth iteration of its open source Llama model, competition for OpenAI and its ChatGPT model to dominate the field of artificial intelligence has grown in the last year.

In addition to training new models, major AI development companies are investing billions of dollars in new data centers to accommodate new iterations. In January 2025, Meta announced the construction of a 2GW data center with more than 1.3 million Nvidia AI graphics processing units.

"This will be a decisive year for AI. By 2025, I expect Meta AI to be the leading assistant serving more than a billion people... To achieve this, Meta is building a data center of more than 2GW, which is large enough to cover a large portion of Manhattan."

Description of the data map coverage. Source: Mark Zuckerberg

His statement comes after the $500 billion Stargate project, which will be led by OpenAI and SoftBank to make a massive investment in artificial intelligence development, with Microsoft and Oracle as equity partners.

In this competition, AI companies are looking for a broader and more diverse data source to train their AI models — and have taken a questionable approach to getting the data they need. To remain competitive with OpenAI when developing the Llama 3 model, Meta has collected thousands of pirated books from the LibGen website. According to court documents in the case against Meta, Llama developers gather data from pirated books because it seems “too expensive” to obtain authorization from sources like Scribd.

Time is another obvious motivation to use pirated works. “It took them over four weeks to deliver the data,” one engineer wrote about the services they could purchase a book license through.

This practice is not limited to Meta. OpenAI is also accused of mining data from pirated works hosted on LibGen.

Rossow recommends that “to ensure lasting impact – beyond short-term profits”, Meta is best to “prioritize investment in advanced data collection, rigorous audits, and the implementation of privacy protection and encryption-based technologies.”

By focusing on transparency and responsible practices, “no matter how the platform portfolio changes, Meta can continue to truly enhance AI capabilities, rebuild and cultivate long-term user trust, and adapt to evolving legal and ethical standards.”

What does the FTC ruling mean?

Currently, tech companies are facing lawsuits from all parties, facing allegations of privacy violations, copyright violations and stifling competition. The unresolved major cases faced by companies such as Google, Amazon and Meta will determine how and whether these companies can continue to grow as they did before, while also determining guardrails for the development of artificial intelligence.

Rosso said the current antitrust case against Meta could determine how the court interprets antitrust laws against tech companies, covering technology mergers, data use and market competition. This also shows that when it comes to issues of stifling competition, the court is "willing to split the tech group" while "morely coordinating current precedents with the laws of cyberspace."