Meta's AI Push Amidst 20 Million User Loss

Meta’s business-AI suite now powers 10 million conversations a week, the company announced on Thursday. This figure comes from Meta’s internal metrics for tools that let advertisers and brands automate chat, generate copy, and route support tickets.

The same earnings call revealed a 20 million-user drop in Meta’s ‘Family daily active people’ metric – the combined tally of Facebook, Instagram, WhatsApp, and Messenger users – compared with the prior three-month period. Meta’s decision to pump billions into AI comes despite millions of users seemingly abandoning its platforms.

AI usage spikes

The 10 million weekly conversations figure represents a concrete usage metric that the company can monetize through its ad-driven revenue model. Over 8 billion advertisers have used at least one of Meta’s generative-AI products. This usage rate indicates that Meta’s AI tools are resonating with advertisers, who see the value in automating customer interactions.

Meta’s business-AI stack rests on its Llama family of large language models, fine-tuned on proprietary social data. The architecture uses on-device inference for privacy-sensitive tasks and cloud-based serving for high-throughput workloads. By leveraging its Llama models, Meta can provide more accurate and efficient AI-powered services to its advertisers.

User decline fuels new AI spending wave

In the same call, Meta pledged to pour ‘billions more’ into AI research and productization this year. The spend will target both foundational model training and the integration of AI into core social experiences. Investors see the AI push as a bid to extract more value from a shrinking advertiser pool.

The 20 million-user decline in Meta’s Family group raises questions about the company’s strategy. Meta attributes this fall to “internet disruptions in Iran, as well as a restriction on access to WhatsApp in Russia.” However, this explanation may not fully account for the decline, and investors are likely to scrutinize Meta’s user metrics closely in future quarters.

Niche AI startups challenge Meta

Smaller players like Polymorph and HippoMod are carving out specialized use cases. Polymorph markets an AI-driven personalization engine, while HippoMod offers AI-powered comment moderation for social media accounts. These startups are targeting specific pain points in the market, and their success could potentially erode Meta’s market share.

Polymorph’s AI personalization engine is designed to help consumer app developers increase user engagement. By providing personalized messages and offers, developers can boost user retention and revenue. HippoMod’s AI-powered comment moderation, on the other hand, helps social media accounts manage their online presence and reduce the risk of harassment or spam.

History of AI Investments

Meta’s AI investments are not a new phenomenon. The company has been pouring resources into AI research and development for years, with a focus on applying AI to improve user experiences and advertising effectiveness. In 2019, Meta announced the launch of its AI Research Lab, which aims to develop and apply AI to solve complex problems.

Meta’s AI investments have also led to the development of various AI-powered tools and features, such as AI-powered chatbots and AI-driven content generation. These tools have been designed to help advertisers and developers automate tasks, improve user engagement, and drive revenue growth.

Broader industry context

The AI market is rapidly evolving, with new applications and use cases emerging every quarter. According to a recent report, the global AI market is expected to reach $190 billion by 2025, growing at a CAGR of 38%. This growth is driven by increasing demand for AI-powered solutions in industries such as customer service, marketing, and healthcare.

Meta’s AI push is part of a larger trend towards AI adoption in the tech industry. Other companies, such as Google and Amazon, are also investing heavily in AI research and development. As the AI market continues to grow, Meta’s investments in AI are likely to pay off, but the company must navigate the challenges of a rapidly changing market.

Technical Mechanics

Meta’s business-AI stack is built on its Llama family of large language models. These models are fine-tuned on proprietary social data and use a combination of on-device inference and cloud-based serving to provide more accurate and efficient AI-powered services.

The use of on-device inference for privacy-sensitive tasks allows Meta to protect user data while still providing AI-powered services. Cloud-based serving, on the other hand, enables Meta to handle high-throughput workloads and provide more scalable AI-powered services.

Downstream implications

The next earnings quarter will reveal whether Meta’s AI investment translates into higher ad-spend per advertiser. Track the ‘average revenue per user’ metric for the Family group and the quarterly update on AI-related operating expenses. If Meta’s AI push leads to increased ad revenue, it could justify the company’s significant investments in AI research and development.

However, if the user decline continues, Meta may face increased pressure to demonstrate the effectiveness of its AI investments. The company’s ability to balance user growth with AI-powered revenue growth will be a key metric to watch in future quarters.

What to watch

The coming quarters will be crucial in determining the success of Meta’s AI push. Investors will be watching closely to see if the company’s AI investments lead to increased revenue and user engagement. As the AI market continues to evolve, Meta must navigate the challenges of a rapidly changing landscape while executing on its AI strategy.

In the short term, investors will be monitoring Meta’s user metrics and AI-related operating expenses. In the long term, they will be watching to see if Meta’s AI investments lead to sustainable revenue growth and increased competitiveness in the market.

Conclusion

Meta’s AI push is a strategic move to drive revenue growth and increase competitiveness in the market. While the company’s user decline raises questions about its strategy, its AI investments have the potential to pay off in the long term. As the AI market continues to evolve, Meta must navigate the challenges of a rapidly changing landscape while executing on its AI strategy.

Updates

• 2026-05-01 — Apple’s R&D spending hits new record as AI investment ramps up (source)