Meta has officially launched its new generative AI assistant, Meta AI, in a standalone mobile app for iOS and Android, just weeks after the company began rolling out artificial intelligence features on Instagram and WhatsApp. Built on the Llama 4 large language model, the assistant offers users a personalized, adaptive experience, with the ability to learn individual preferences, retains contextual memory across interactions, and adjusts to evolving user behavior.
Meta AI – Mark Zuckerberg
Meta’s arrival on an already competitive market.
The release marks a significant step in Meta’s broader artificial intelligence strategy, as the company races to catch up in an increasingly crowded market of AI-powered applications. Despite entering the space later than competitors like OpenAI’s ChatGPT, Google’s Gemini, and Microsoft’s Copilot, Meta has already invested tens of billions of dollars into generative AI, signaling a long-term commitment to the technology.
With Meta AI now accessible through a dedicated mobile app, the social media giant aims to reach users where they already spend much of their time. Integration across the Meta ecosystem—Facebook, Instagram, WhatsApp, and Messenger—means that close to one billion people have already interacted with Meta AI in some form, according to internal company estimates.
What sets Meta’s approach apart is a deliberate focus on the social dimension of AI. The app features a “Discover” feed, designed to let users share prompts, ideas, and practical use cases with friends and their wider social circles. This feature taps into Meta’s core strength as a social networking platform, aiming to make AI not just a solitary tool, but a collaborative and community-driven experience.
Data collection amid regulatory scrutiny in Europe.
At the same time, Meta is leveraging its vast pool of user data—collected over years from its global user base—to enhance the performance and contextual intelligence of Meta AI. On April 14, the company announced it would resume a controversial initiative to collect public user data to train its AI models. This includes information from Facebook, Instagram, and WhatsApp profiles, unless users specifically opt out. Even users in Europe, where data privacy laws are among the strictest in the world, are included in this effort unless they file a formal request for exclusion.
This approach, however, has drawn scrutiny from European regulators. On Wednesday, Meta issued a warning that users in Europe could soon face a “materially worse” experience on its platforms as a result of recent enforcement action by the European Commission. At the center of the dispute is Meta’s “consent or pay” model, introduced last year, which gives users a binary choice: either agree to let Meta aggregate and use their data across its services, or pay a monthly subscription fee for an ad-free experience.
Last week, the European Commission ruled that this model does not comply with the Digital Markets Act (DMA), which aims to promote fair competition and protect consumer rights in the digital economy. The Commission fined Meta €200 million, stating that the company does not offer users genuine and freely given consent when it comes to data usage.
In response, Meta stated in its quarterly earnings report that it anticipates needing to modify its business model in Europe in light of the feedback from the Commission. These changes, the company warned, “could result in a materially worse user experience for European users and a significant impact” on Meta’s revenue from the region.
An important strategic investment for the company after a strong yearly start.
Despite these regulatory setbacks, Meta CEO Mark Zuckerberg remains optimistic about the company’s trajectory in AI. Speaking on a conference call with investors, he described the opportunities in artificial intelligence as “staggering,” and confirmed that Meta will continue to aggressively invest in AI development.
To support this ambition, Meta has significantly increased its capital expenditure plans. After initially projecting to spend between $60 billion and $65 billion in 2025, the company has now raised that range to between $64 billion and $72 billion. These investments will primarily fund the expansion of data centers, server infrastructure, and networking equipment—critical components for scaling AI models and services.
The 2025 capital budget represents a sharp rise from the $39 billion Meta spent in 2024, underlining the company’s determination to remain competitive in the next generation of tech innovation. The announcement came after Meta reported stronger-than-expected quarterly earnings, further reinforcing investor confidence in its long-term AI strategy.
As Meta moves deeper into the generative AI space while simultaneously navigating a shifting regulatory environment, the company finds itself at a critical juncture—balancing innovation with compliance, and user growth with data ethics.