Meta CEO Mark Zuckerberg used the company's Q2 2026 earnings call to set out plans for a major push into personal AI agents that can act autonomously on users' behalf, covering areas such as health, finances and relationships. He described these tools as "the foundation for our next wave of products and revenue lines", positioning Meta to compete not only with coding- and enterprise-focused rivals such as Anthropic and OpenAI, but also with Google, whose Gemini Spark tool already offers similar personal-agent features. Zuckerberg said Meta would share further details "soon", while acknowledging its AI efforts still trail the biggest labs and that, unlike Google or Microsoft, it lacks deep access to users' email and documents — a gap that, combined with public wariness over Meta's smart glasses, could make adoption harder.

Zuckerberg also highlighted existing traction with Meta's business agents on WhatsApp and Messenger, used by more than a million businesses weekly, with the feature now expanding to Instagram, which he said has surpassed two billion daily active users. The personal-agent push follows last quarter's launch of Meta's Muse Spark AI model, recently updated to improve coding capabilities, alongside a company-wide restructuring that has shifted roughly 7,000 staff onto AI projects even as around 8,000 jobs were cut in May. Meta expects to spend $130–145 billion on capital expenditure in 2026, underscoring heavy investment in compute, including a newly announced 1-gigawatt data centre campus built in partnership with BlackRock.