ByteDance funding AI drug discovery content

The core issue is not that ByteDance is investing broadly, it is that its biggest cash engine is now being asked to fund several very different bets at once. TikTok and Douyin still throw off massive revenue, with ByteDance reaching an estimated $186B in 2025, but profit fell more than 70% as AI infrastructure spending surged. At the same time, the company is carrying mature content apps, making large outside investments like Seres, and separating an AI drug discovery business that follows a completely different operating model from ads, commerce, and creator tools.1 2 3

• The AI ramp is large enough to change how the whole company gets valued. ByteDance is considering as much as $70B of 2026 capex after roughly ¥160B, about $23B, of AI infrastructure spend in 2025. That pushes ByteDance closer to the capital intensity of a hyperscaler than a consumer app company.1 4
• The drug discovery spinoff shows management knows some bets do not fit cleanly inside the core machine. Anew Labs was spun out, then raised $290M at a $1.5B valuation, with ByteDance retaining 56%. That structure separates biotech style R&D timelines from the fast feedback loops of ads, content ranking, and creator monetization.1 5
• The contrast with peers is sharp. ByteDance is already juggling short video, live commerce, search, and AI apps inside one stack, while domestic social rivals like Xiaohongshu remain far narrower. That makes every extra non core capital commitment more noticeable because the base business is already complex and regulation heavy.1 6

Going forward, ByteDance is likely to look less like a focused social media winner and more like a self funded Chinese compute platform with several attached businesses. The companies that compound best from here will be the ones that turn AI spend into products and revenue quickly, and ByteDance is clearly trying to make that transition before its legacy apps mature further.1 2 4