Meta looks to Google’s Gemini for ads while facing EU charges

• Meta tests Google’s Gemini to boost ads, suggesting its AI challenges.
• Also launches Vibes and reorganised AI under Superintelligence Labs.

Meta is weighing whether to bring in outside help for its advertising technology. According to The Informationthe company’s staff have held early talks with Google Cloud about using its Gemini models, as well as Gemma, an open-source option, to fine-tune ad targeting. The idea under discussion is to train Google’s systems with Meta’s advertising data to sharpen how ads are matched to users.

The possibility of Meta turning to a competitor highlights one of the company’s internal challenges. Despite years of heavy investment, Meta has struggled to scale its in-house AI. It has poured billions of dollars into research and computing power and hired top researchers, but its results have lagged. The company’s open-source Llama models, once seen as a rival to OpenAI’s GPT series, have faced technical setbacks and a mixed reception from developers. Staff departures in recent months have only sharpened concerns about whether Meta can keep pace with rivals in applied AI.

Both Meta and Google earn the majority of their revenue from advertising, making the talks notable given the direct competition between the two. In their latest quarterly earnings figures, both companies credited AI systems with helping drive higher returns from ads, whether by improving targeting or automating creative output. If Meta were to adopt Google’s technology, it would amount to borrowing tools from one of its biggest rivals operating in the same market.

The talks do not appear to be limited to ads. Last month, The Information reported that Meta had also considered working with either Google or OpenAI to power features in Meta AI, its chatbot. That included conversational responses for search queries and new functions for Instagram and Facebook. The willingness to work with rivals suggests Meta sees value in bringing outside technology into its social platforms, even if it has long-promoted its own AI’s work.

At the same time, Meta continues to push forward with its own AI projects. On Thursday, it launched Vibes, a feed of AI-generated videos. The tool allows users to make clips from scratch, remix others, or add effects and music before sharing them to the Vibes feed. Content can be posted to Instagram and Facebook stories and reels. The launch is part of Meta’s effort to build new experiences around generative AI for users, not just tools for advertisers.

The rollout comes as Meta reorganises its AI work under a new unit called Superintelligence Labs. The division was formed in June, and is tasked with unifying projects and to develop new products that could generate revenue, and comes after setbacks with Llama 4. Among the new division’s goals are expanding the Meta AI app, creating ad tools that turn images into videos, and linking AI to hardware devices such as smart glasses. Meta reported nearly $165 billion in revenue last year, and is looking for ways to ensure AI becomes a source of future income rather than a research expense.

While the company experiments with new products, it is also facing mounting regulatory pressure. Bloomberg reported that Meta will soon receive a charge sheet from the European Union, accusing it of failing to adequately police illegal content. According to people familiar with the plans, the European Commission is preparing preliminary findings that Facebook and Instagram lack an effective system for users to flag and remove harmful posts.

If confirmed, Meta could face fines of up to 6% of its annual global sales under the EU’s Digital Services Act, which could amount to many billions of dollars. The company would have the chance to respond before any final ruling is issued. “We disagree with any suggestion we have breached the DSA,” Meta spokesperson Ben Walters said, adding that the company continues to negotiate with regulators.

The DSA requires platforms and search engines with more than 45 million monthly users in Europe to prevent the spread of illegal or harmful material. Meta is currently the subject of two EU probes – one on illegal content and disinformation, and another on how it protects minors. Other large firms under investigation include TikTok, Elon Musk’s X, Temu, and AliExpress.

The law has become a flash-point between Europe and the United States. No fines have been imposed yet, but American firms dominate the list of targets. US President Donald Trump has criticised action under the DSA as political moves aimed at restricting American tech companies, and has accused the EU of limiting free speech.

The tension extends beyond the DSA. Apple this week asked the EU to roll back parts of the Digital Markets Act, another regulatory rule book. Apple argued the rules weaken iPhone performance and expose users to more scams and malware. Brussels rejected the appeal, saying there are no plans to revise the laws.

For Meta, the mix of technical hurdles, new product bets, and regulatory risks shows how much is at stake in its AI strategy. The company is trying to prove it can still shape the future of social platforms while having to defend itself in one of the world’s toughest regulatory environments.

Want to experience the full spectrum of enterprise technology innovation? Join TechEx in Amsterdam, California, and London. Covering AI, Big Data, Cyber Security, IoT, Digital Transformation, Intelligent Automation, Edge Computing, and Data Centres, TechEx brings together global leaders to share real-world use cases and in-depth insights. Click here for more information.

TechHQ is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.