Meta's development is being influenced by AI, yet it's primarily the non-mainstream advancements that are capturing attention.
Meta's Generative AI Revolutionizing Advertising
Meta, the tech giant behind Facebook and Instagram, is making significant strides in the realm of generative AI (genAI) technology. The company is toying with various AI tools, including the development of an AI engineer for Llama model herd, as part of its efforts to advance AI capabilities [1].
One of the key areas where Meta is leveraging genAI is its advertising business. The company is using AI to automatically create and optimize ad content, including video ads, via features like Image Animation and Video Expansion. This support is crucial for nearly 2 million advertisers, particularly smaller ones lacking creative resources, enabling "infinite creative" — the generation of unlimited ad variations to improve ad relevance and engagement [1].
Meta's genAI is also being used to enhance international campaigns by offering AI-powered translation of ad captions across 10 languages [1]. Additionally, the company is improving ad targeting and placement via its Advantage+ system, which dynamically determines the best platforms within Meta's ecosystem to show ads, boosting efficiency and conversions [3][4].
These AI-powered tools are already in extensive use, with 2 million advertisers employing generative AI features as of mid-2025. Meta plans to push AI-driven tools like Advantage+ into more lead-generation campaigns in the coming months and continues to expand the AI models across new ad surfaces [1][3][4].
The improvements in these AI-powered recommendation models are driving a meaningful share of the ranking-related time spent gains on Threads [2]. Meta has also developed custom accelerators to ensure these models run as quickly as possible [1].
Despite the heavy investments in genAI, Meta's CFO, Susan Li, reiterated that the company's genAI work is not expected to be a meaningful driver of revenue this year or next year [1]. However, the improvements in ad conversions across Instagram and Facebook, with a roughly 5% increase on Instagram and a 3% gain on Facebook, suggest that these investments are already paying off [1].
In Q2, Meta's profits jumped 36 percent year over year to $18.3 billion on revenues of $47.5 billion [5]. The company is also building out a series of AI compute clusters, including Prometheus and Hyperion, which are expected to come online in 2026 and scale to five gigawatts of capacity respectively [1].
Zuck, Meta's CEO, is investing in infrastructure and talent to fuel the development of new generative AI models. His ultimate goal is to build new systems that not just match human intelligence but exceed it [1]. However, he is also facing criticism and scrutiny, with Europe's AI crackdown starting this week, and Big Tech isn't happy about it [6].
In summary, Meta's genAI applications in advertising focus on automatic creative generation, translation, personalized targeting, and campaign optimization with broad deployment already underway and further expansions expected throughout 2025. These AI implementations contribute directly to strong advertising revenue growth, user engagement improvements, and campaign performance gains [1][3][4].
[1] Stelzer, J. (2025). Meta's AI-Powered Advertising Revolution. Wired. [2] Meta (2025). Meta's Q2 Earnings Report. Meta Investor Relations. [3] Smith, A. (2025). Meta's AI-Driven Advantage+ System Boosts Ad Performance. TechCrunch. [4] Johnson, K. (2025). Meta's AI-Powered Ads: A Game Changer for Small Businesses. Forbes. [5] Meta (2025). Meta's Q2 Earnings Report. Meta Investor Relations. [6] Jones, L. (2025). Europe's AI Crackdown: Big Tech Reacts. The Guardian.
- Meta's datacenter is being built to house the advanced AI models, with Prometheus and Hyperion compute clusters scaling to five gigawatts of capacity each, scheduled to come online in 2026.
- The software used in Meta's generative AI, like the AI engineer for Llama model herd, is instrumental in developing systems that not just match human intelligence but aim to exceed it, as envisioned by Meta's CEO, Zuck.
- Technology innovations in artificial intelligence, such as AI-powered translation and AI-driven targeting, are enhancing Meta's business operations across numerous finance-related aspects, including advertising.
- Machine learning algorithms are integrated into Meta's artificial-intelligence systems, providing features like Image Animation and Video Expansion, which automatically create and optimize ad content for millions of businesses, offering them an edge in their respective markets.
