China's three largest tech companies are undergoing dramatic organizational overhauls to prioritize artificial intelligence, with ByteDance, Alibaba, and Tencent collectively spending hundreds of billions of yuan on AI infrastructure and talent in 2026. ByteDance has merged its Feishu productivity platform into its Doubao AI division, signaling a strategic shift to make AI independent rather than attached to legacy business units. This restructuring reflects a broader pattern among major tech giants who are treating AI as a standalone competitive moat rather than an add-on feature.

The scale of investment tells the story of how seriously these companies are taking the AI race. ByteDance raised its capital expenditure from 160 billion yuan in 2025 to 200 billion yuan in 2026, with some reports suggesting the figure could reach as high as 70 billion U.S. dollars. Alibaba announced a 380 billion yuan investment over three years for cloud and AI hardware infrastructure, with 190 billion yuan already deployed by mid-2026. Tencent's capital expenditure surged to 84.72 billion yuan in the first half of 2026 alone, exceeding its entire 2025 spending of 79.2 billion yuan.

Why Are China's Tech Giants Spending So Much on AI Infrastructure?

The answer lies in the competitive urgency that has gripped the industry. After two years of exploratory AI projects, the technology has moved from experimental to existential for these companies. Each giant recognizes that whoever builds the most robust AI infrastructure and attracts top talent will establish an unbeatable advantage in the coming decade. The organizational restructuring, more than the capital spending alone, reveals how seriously executives view this transition.

ByteDance's decision to merge Feishu into Doubao exemplifies this strategic thinking. Rather than maintaining AI as a separate research division, the company is consolidating its resources to create a unified AI platform. This mirrors similar moves by Tencent, which reorganized its AI departments under a single leadership structure. In December 2025, Tencent appointed Yao Shunyu, a former OpenAI researcher, as Chief AI Scientist, reporting directly to President Martin Lau. Under his leadership, Tencent established three core departments: AI Infra Department, AI Data Department, and Data Computing Platform Department.

How Are These Companies Restructuring Their AI Operations?

• Consolidation Strategy: ByteDance merged Feishu into Doubao to create a unified AI business unit, eliminating the separation between productivity tools and AI models.
• Centralized Leadership: Tencent appointed a former OpenAI researcher as Chief AI Scientist with direct authority over multiple AI departments, streamlining decision-making and resource allocation.
• Talent Acquisition: All three companies are competing aggressively for top AI researchers, with ByteDance reportedly offering compensation packages worth nearly 100 million yuan annually to poach talent from competitors like DeepSeek.

The talent war reflects the stakes involved. In 2023, top AI researchers commanded annual salaries in the millions of yuan. By 2026, ByteDance poached Guo Daya from DeepSeek with compensation reportedly approaching 100 million yuan annually. This escalation shows how desperately these companies are competing for the expertise needed to build next-generation AI systems.

Tencent's reorganization has already yielded results. Less than four months after Yao Shunyu joined the company, Tencent launched Hunyuan Hy3, a large language model that achieves performance comparable to or exceeding some flagship models while using smaller activation parameters, meaning it requires less computing power to run. The company's AI products, including WorkBuddy, a desktop office agent, now rank first among similar domestic products in monthly visits.

The financial pressure is real, even for companies with substantial cash reserves. Alibaba's free cash flow turned negative in the second quarter of 2026, with a net outflow of 44.67 billion yuan. Despite holding approximately 474.5 billion yuan in cash and liquid investments, Alibaba chose to raise 80 billion Hong Kong dollars through a new share placement. This decision signals that the company views AI investment not merely as a capital expenditure but as a strategic necessity that requires external validation from investors.

"The company's AI investment is mainly focused on AI infrastructure. In the worst case, these infrastructures can also be leased out through Tencent Cloud, and the company will dynamically adjust the investment scale," explained Martin Lau, President of Tencent.

Martin Lau, President of Tencent

Tencent's approach differs slightly from its competitors. While Alibaba and ByteDance are investing aggressively with less flexibility, Tencent has built optionality into its strategy. The company can lease excess computing capacity through Tencent Cloud, converting infrastructure spending into potential revenue streams. This hedging strategy reflects Tencent's more cautious approach, though the company's free cash flow also turned negative in the second quarter of 2026, reaching negative 13.8 billion yuan.

The organizational restructuring also addresses a critical challenge: making AI profitable. ByteDance's net profit under International Financial Reporting Standards (IFRS) accounting plummeted approximately 70 percent in 2025, though company executives disputed whether this metric accurately reflected operational performance. Profits are thinning across the board, yet valuations continue to climb. ByteDance's reference price in secondary market share transactions rose from 480 billion U.S. dollars in November 2025 to 550 billion U.S. dollars by February 2026, suggesting investors believe these companies can convert AI infrastructure into sustainable revenue.

The competitive dynamics have shifted fundamentally. A few years ago, major tech giants could afford to experiment with multiple AI approaches simultaneously, a strategy known as "horse racing." By 2026, that luxury no longer exists. The sunk costs are too high, and the competitive window is too narrow. Every company must commit fully to its chosen path, knowing that hesitation could mean falling permanently behind. The organizational restructuring at ByteDance, Alibaba, and Tencent signals that the exploratory phase of AI development has ended and the winner-take-most competition has begun.

" }