Meta's $18 billion settlement may unlock new AI push, Morgan Stanley says

The settlement ends a case that went to trial in August in Oakland federal court and had carried a potential penalty Meta itself put as high as $1.4 trillion in court filings. Meta agreed to platform changes for minors, including a two-hour daily use cap, blocking apps for children at night, removing certain cosmetic filters, and tighter age verification, and expects to record about $10 billion in legal expense in the third quarter — a charge not included in the guidance it gave during its July earnings report.

Morgan Stanley draws a comparison to Google's resolution of a major antitrust overhang last year, which was followed by successful launches including Gemini 3 and expanded AI Overviews. On that reading, the settlement removes a legal risk ahead of Meta's reported plans for a consumer AI agent called Hatch inside WhatsApp and Instagram in early September. Meta's aggressive AI investment plans have not been scaled back, and it still expects $130 billion to $145 billion in capital expenditures in 2026, much of it for AI infrastructure.

The bear case sits on timing. Needham analysts said the payment schedule could not look worse because Meta already plans to spend as much as $145 billion on 2026 capital expenditures for the AI race, and the roughly $10 billion legal expense adds another major cost on top. How much Meta ultimately pays hinges on whether YouTube and TikTok settle and adopt similar youth-safety measures, since about $5.3 billion of the $18 billion is conditional on that. For Meta's product and infrastructure teams, the outcome appears to preserve the room to keep launching; for its finance team, the near-term burden is the legal charge landing alongside already-elevated spending.