Global user engagement on Instagram surged by a double-digit percentage year-over-year during the second quarter, driven primarily by algorithmic upgrades to feed and Reels recommendations, according to Meta CEO Mark Zuckerberg. Speaking during an earnings call, Zuckerberg reported that sessions on the platform rose by 15 basis points, fueled by enhanced personalization and faster data inference architectures.
Algorithmic Upgrades and Increased User Engagement
According to Meta, the surge in time spent on Instagram stems directly from technical improvements in how content is served to users. Zuckerberg stated during the second-quarter earnings call that recommendation systems now combine faster inference with a novel architecture based on a broader user history to refine content predictions. These adjustments resulted in distinct increases in content resharing and overall dwell time. Meta Chief Financial Officer Susan Li joined the call to outline the company’s financial metrics alongside these engagement updates.
The platform’s recommendation engine operates by automatically feeding every public reel and post into a large language model to analyze its topic and tone. Zuckerberg described this ingestion process as a core building block for personalization, noting that the company plans to expand these recommendation frameworks into Facebook interfaces as well.
Regulatory Scrutiny and Youth Safety Lawsuits
The algorithmic push coincides with heightened legal and regulatory pressure regarding youth safety on social media. More than two dozen US-Bundesstaaten have filed lawsuits against Meta, alleging that the company deliberately designs its products to foster addictive behaviors among children and teenagers. During the earnings call, CFO Susan Li stated that Meta anticipates continued regulatory scrutiny across multiple markets regarding youth-related issues.
Li also noted that a series of youth-related legal proceedings are scheduled for the year, which could result in financial liabilities. According to the company’s financial disclosures, Meta recorded $2.40 billion in expenses related to legal proceedings.
Privacy Controversies Surrounding AI Image Tools
Meta’s data ingestion practices have drawn separate regulatory and user backlash. In early July, the company introduced an AI image generation tool called Muse Image, which allowed users to generate artificial images using photos from public Instagram accounts belonging to individuals over 18. Eligible public accounts were enrolled by default without advance notice.
The rollout triggered an immediate privacy controversy, prompting Meta to shut down the image generation feature within a week of its launch. Users retain the ability to block Meta from utilizing visual content by navigating to the sharing and reuse section within Instagram settings, though no equivalent opt-out mechanism exists for text, comments, or audio files.
Related reading