Listen to this article in summarized format
Loading...
××
Subscribe to
Unlock AI Briefing and Premium Content
New Year Offer 24 Hours Left
Subscribe NowAlready a member? Sign In
What's Included
- Exclusive Stories
- Daily ePaper Access
- Smart Market Tools
- Curated Investment Ideas
- Ad-lite Experience
Subscription
Artificial intelligence (AI) is driving gains in engagement and advertising at Meta Platforms, with the company reporting higher video consumption, improved ad conversions, and wider adoption of AI tools across its apps in the first quarter of FY26.
On Instagram, “ranking improvements… drove a 10% lift in Reels time spent,” Meta chief executive Mark Zuckerberg said, while on Facebook, total video watch time rose more than 8% globally, including a 9% increase in the US and Canada.
Zuckerberg added that more than 3.5 billion people now use at least one of the company’s apps daily, underscoring the scale at which these systems operate.
The gains come as Meta expands the use of AI-powered recommendation systems to surface more relevant content to users.
Also Read: Meta layoffs: More cuts possible after 10% workforce reduction on May 20
AI reshapes content discovery
Meta said improvements in its recommendation infrastructure are enabling faster and more personalised content delivery. The company has increased the depth of data used to train its models, including longer user interaction histories that describe how users engage with content.
“We’ve significantly increased the speed with which our ranking models index new posts… enabling us to recommend them sooner after they are published,” Zuckerberg said.
As a result, same-day posts now account for more than 30% of recommended Reels across Instagram and Facebook, more than double the level a year earlier.
The company is also using AI to expand content accessibility. “Over half a billion users on each of Facebook and Instagram are now watching AI-translated videos weekly,” he added, referring to tools that automatically translate and dub video content into local languages.
Meta added that it is continuing to refine large language model-based systems that allow users to provide more granular feedback on content preferences, helping improve personalisation over time.
“We’re continuing to improve our data infrastructure so our models can train on more data. We’re also focussed on making recommendations more personalised and relevant to each user’s interests,” said Susan Li, Meta’s chief financial officer.
“Users with more concentrated interests may see more of that specific content, while those with broader interests will see a wider range of topics.”
AI-led ad gains
Advances in AI are also contributing to improvements in Meta’s core advertising business, which accounts for the majority of its revenue.
Average price per ad increased 12% on-year, per the company’s financial disclosures.
Li added that more than 8 million advertisers now use at least one of the company’s generative AI ad tools, up from 4 million at the end of 2024. These include video generation, image animation, and automated creative optimisation across formats such as Reels, Feed, and Stories.
Li said that advertisers using its video generation feature saw a 3% increase in conversion rates compared with those that did not use the tool.
The company also reported broader gains from upgrades to its ad delivery systems. Enhancements to its Lattice and GEM model architectures drove “more than a 6% increase in the conversion rate for landing page view ads” during the quarter. These are Meta’s internal AI systems used to decide which ads to show, to whom, and when.
Meta said its Adaptive Ranking Model, an AI-based system that routes ad requests to the most effective and efficient models where conversion probability is higher, contributed to a 1.6% increase in conversion rates across its major platforms.
In addition, Meta’s AI-powered business assistant for advertisers has now been fully rolled out, providing campaign recommendations and resolving account issues. The company said common advertiser issues are now being resolved “at a 20% higher rate” compared with earlier testing phases.
“There is still a lot of room to continue improving recommendations over the rest of the year, and we expect we’ll be able to do that to drive additional engagement on both Facebook and Instagram,” Li added.
Scaling AI infra
Meta said it is continuing to invest heavily in infrastructure to support its AI roadmap, including servers, data centres, and network capacity. Capital expenditure for the quarter stood at $19.8 billion, reflecting increased investment in compute resources required for training and deploying advanced models.
Meta reported a first-quarter operating income of $22.9 billion, representing a 41% operating margin, while its net income stood at $26.8 billion.
The company said it expects continued improvements in recommendation quality and ad performance as it scales its AI systems further across its platforms.
“There’s going to be massive upside in delivering more intelligence and capabilities. This is one of the most exciting times in the industry,” Zuckerberg said, and added: “AI agents are the product vehicle for bringing those capabilities to people, and this year will be a key period for establishing that,” Zuckerberg added.
On Instagram, “ranking improvements… drove a 10% lift in Reels time spent,” Meta chief executive Mark Zuckerberg said, while on Facebook, total video watch time rose more than 8% globally, including a 9% increase in the US and Canada.
Zuckerberg added that more than 3.5 billion people now use at least one of the company’s apps daily, underscoring the scale at which these systems operate.
The gains come as Meta expands the use of AI-powered recommendation systems to surface more relevant content to users.
Also Read: Meta layoffs: More cuts possible after 10% workforce reduction on May 20
AI reshapes content discovery
Meta said improvements in its recommendation infrastructure are enabling faster and more personalised content delivery. The company has increased the depth of data used to train its models, including longer user interaction histories that describe how users engage with content.
“We’ve significantly increased the speed with which our ranking models index new posts… enabling us to recommend them sooner after they are published,” Zuckerberg said.
As a result, same-day posts now account for more than 30% of recommended Reels across Instagram and Facebook, more than double the level a year earlier.
The company is also using AI to expand content accessibility. “Over half a billion users on each of Facebook and Instagram are now watching AI-translated videos weekly,” he added, referring to tools that automatically translate and dub video content into local languages.
Meta added that it is continuing to refine large language model-based systems that allow users to provide more granular feedback on content preferences, helping improve personalisation over time.
“We’re continuing to improve our data infrastructure so our models can train on more data. We’re also focussed on making recommendations more personalised and relevant to each user’s interests,” said Susan Li, Meta’s chief financial officer.
“Users with more concentrated interests may see more of that specific content, while those with broader interests will see a wider range of topics.”
AI-led ad gains
Advances in AI are also contributing to improvements in Meta’s core advertising business, which accounts for the majority of its revenue.
Average price per ad increased 12% on-year, per the company’s financial disclosures.
Li added that more than 8 million advertisers now use at least one of the company’s generative AI ad tools, up from 4 million at the end of 2024. These include video generation, image animation, and automated creative optimisation across formats such as Reels, Feed, and Stories.
Li said that advertisers using its video generation feature saw a 3% increase in conversion rates compared with those that did not use the tool.
The company also reported broader gains from upgrades to its ad delivery systems. Enhancements to its Lattice and GEM model architectures drove “more than a 6% increase in the conversion rate for landing page view ads” during the quarter. These are Meta’s internal AI systems used to decide which ads to show, to whom, and when.
Meta said its Adaptive Ranking Model, an AI-based system that routes ad requests to the most effective and efficient models where conversion probability is higher, contributed to a 1.6% increase in conversion rates across its major platforms.
In addition, Meta’s AI-powered business assistant for advertisers has now been fully rolled out, providing campaign recommendations and resolving account issues. The company said common advertiser issues are now being resolved “at a 20% higher rate” compared with earlier testing phases.
“There is still a lot of room to continue improving recommendations over the rest of the year, and we expect we’ll be able to do that to drive additional engagement on both Facebook and Instagram,” Li added.
Scaling AI infra
Meta said it is continuing to invest heavily in infrastructure to support its AI roadmap, including servers, data centres, and network capacity. Capital expenditure for the quarter stood at $19.8 billion, reflecting increased investment in compute resources required for training and deploying advanced models.
Meta reported a first-quarter operating income of $22.9 billion, representing a 41% operating margin, while its net income stood at $26.8 billion.
The company said it expects continued improvements in recommendation quality and ad performance as it scales its AI systems further across its platforms.
“There’s going to be massive upside in delivering more intelligence and capabilities. This is one of the most exciting times in the industry,” Zuckerberg said, and added: “AI agents are the product vehicle for bringing those capabilities to people, and this year will be a key period for establishing that,” Zuckerberg added.
