Tech giant Meta has been given the green light from the European Union’s data regulator to train its artificial intelligence models using publicly shared content across its social media platforms.

Posts and comments from adult users across Meta’s stable of platforms, including Facebook, Instagram, WhatsApp and Messenger, along with questions and queries to the company’s AI assistant, will now be used to improve its AI models, Meta said in an April 14 blog post.

The company said it’s “important for our generative AI models to be trained on a variety of data so they can understand the incredible and diverse nuances and complexities that make up European communities.”

Meta has a green light from data regulators in the EU to train its AI models using publicly shared content on social media. Source: Meta

“That means everything from dialects and colloquialisms, to hyper-local knowledge and the distinct ways different countries use humor and sarcasm on our products,” it added.

However, people’s private messages with friends, family and public data from EU account holders under the age of 18 are still off limits, according to Meta.

People can also opt out of having their data used for AI training through a form that Meta says will be sent in-app, via email and “easy to find, read, and use.”

EU regulators paused tech firms’ AI training plans

Last July, Meta delayed training its AI using public content across its platforms after privacy advocacy group None of Your Business filed complaints in 11 European countries, which saw the Irish Data Protection Commission (IDPC) request a rollout pause until a review was conducted.

The complaints claimed Meta’s privacy policy changes would have allowed the company to use years of personal posts, private images, and online tracking data to train its AI products.  

Meta says it has now received permission from the EU’s data protection regulator, the European Data Protection Commission, that its AI training approach meets legal obligations, and the company continues to engage “constructively with the IDPC.”

“This is how we have been training our generative AI models for other regions since launch,” Meta said.

“We’re following the example set by others, including Google and OpenAI, both of which have already used data from European users to train their AI models.”

Related: EU could fine Elon Musk’s X $1B over illicit content, disinformation

An Irish data regulator opened a cross-border investigation into Google Ireland Limited last September to determine whether the tech giant followed EU data protection laws while developing its AI models.

X faced similar scrutiny and agreed to stop using personal data from users in the EU and European Economic Area last September. Previously, X used this data to train its artificial intelligence chatbot Grok. 

The EU launched its AI Act in August 2024, establishing a legal framework for the technology that included data quality, security and privacy provisions. 

Magazine: XRP win leaves Ripple a ‘bad actor’ with no crypto legal precedent set

The governor of the Northern Mariana Islands, a small Pacific US territory just north of Guam, has killed the legislation that would have allowed one of the territory’s local governments to launch a fully backed US dollar-pegged stablecoin.

In an April 11 letter seen by Cointelegraph, Northern Mariana Islands Governor Arnold Palacios said he vetoed the bill as it “presents several legal issues and may be unconstitutional.”

Palacios’ letter said the bill, which largely dealt with issuing licenses to internet casinos, would regulate an activity that could not “be clearly restricted” to Tinian, a small island forming part of the territory that was hoping to launch a stablecoin.

Tinian, which has just over 2,000 residents and a largely tourism-based economy, is governed by the local government, the Municipality of Tinian and Aguiguan, one of four municipalities in the Commonwealth of the Northern Mariana Islands.

In February, Republican Northern Marianas Senator Jude Hofschneider led the introduction of the bill to amend a local Tinian law to allow internet-only casino licenses, which tacked on a provision allowing the Tinian treasurer to issue, manage and redeem a “Tinian Stable Token.” 

The four-member Tinian delegation to the Marianas legislature passed the bill in a unanimous vote on March 12.

In vetoing the bill, Palacios didn’t comment on the proposed stablecoin, instead taking issue with its aim to police an industry that can cross jurisdictional boundaries, and said the measure lacked “robust enforcement measures to prevent illegal gaming activities.”

A highlighted excerpt of Governer Palacios’ letter noting his reasons for vetoing the stablecoin and internet gambling bill Source: Northern Mariana Islands Governor’s Office

Tinian misses chance at beating Wyoming

The bill’s passage could have seen Tinian’s government be the first US government entity to issue a stablecoin ahead of Wyoming, whose Governor Mark Gordon said in March that the state’s stablecoin could be ready for a launch in July.

The stablecoin was to be known as the Marianas US Dollar (MUSD), which was to be fully backed by cash and US Treasury bills held in reserve by the Tinian Municipal Treasury, according to statements shared with Cointelegraph last month.

Related: The GENIUS stablecoin bill is a CBDC trojan horse — DeFi exec

The Tinian local government chose tech services firm Marianas Rai Corporation, based in the Commonwealth’s capital of Saipan, to exclusively provide the infrastructure to issue and redeem MUSD and develop its ecosystem.

The token was slated to launch on the eCash blockchain, a network that rebranded from Bitcoin Cash ABC in 2021 and is a fork of Bitcoin Cash — a blockchain that split off from Bitcoin in 2017.

The launch of MUSD was meant to coincide with Google’s $1 billion plan announced in April to route fiber-optic subsea cables from the mainland US through Tinian and onto Japan to improve internet connectivity.

Magazine: Elon Musk’s plan to run government on blockchain faces uphill battle 

There is a “huge opportunity” for Ethereum to become a decentralized partner in solving current problems with artificial intelligence platforms, according to a former core Ethereum developer. 

Ethereum’s “biggest mainstream moment is waiting in the wings with AI,” said Eric Connor on X on April 15. 

While AI is on a fast track to reshape almost every aspect of our lives, it is “plagued by black-box models, centralized data silos, and privacy pitfalls,” Connor continued. However, these problems create an opportunity for Ethereum to shine, he said.

Ethereum offers transparency through verifiable smart contracts, decentralization against big tech monopolies, aligned incentives via token economies, and built-in micropayment infrastructure, he added. 

Smart contracts can provide transparent records of AI model training processes and data sources, addressing the “black box” problem.

However, major AI players may resist open models “as they profit from secrecy and control,” he said. Demand for transparency, fairness, and security will only grow, and that’s where Ethereum “offers an alternative path,” he added. 

“Ethereum already has the ethos with openness, collaboration and trust minimization,  things that ethical and accountable AI needs.”

By proactively building the tooling, research and real-world use cases, Ethereum can give AI developers a reason to embrace decentralized approaches, “and that could deliver mainstream adoption far beyond finance,” Connor concluded. 

The next frontier for crypto will be decentralizing AI, Zain Jaffer, co-founder of Vungle, told Cointelegraph earlier this year. 

Connor left the Ethereum community in January amid growing leadership concerns to pursue interests in AI. 

AI agents on Ethereum

Ethereum may also be important for the development of agentic AI — an emerging and experimental technology,  according to a recent post on the Ethereum blog. 

AI agents are software programs that use artificial intelligence to autonomously perform tasks, make decisions, learn from data, and adapt to changes, and they are growing in numbers on Ethereum.

The Ethereum blockchain provides key advantages for AI agents, including access to transparent, real-time blockchain data, true digital asset ownership, and the ability to execute transactions and interact with smart contracts, it noted.

Related: The future of digital self-governance: AI agents in crypto

The post highlighted three notable projects, which were Luna, an autonomous virtual influencer that controls its own onchain wallet; ¡` ×AIXBT, an AI agent providing crypto market analysis; and Botto, a decentralized autonomous artist creating NFTs guided by community voting.

Meanwhile, projects like Bankr and HeyAnon are simplifying blockchain interactions through conversational interfaces, allowing users to manage wallets and execute transactions via simple chat commands.

Magazine: Bitcoin eyes $100K by June, Shaq to settle NFT lawsuit, and more: Hodler’s Digest