AI Is Redistributing Ad Budgets, Not Expanding the Pie — Yet
TREE NEWS reports: Artificial intelligence is already reshaping how advertising budgets flow across China’s internet platforms, but the shift is primarily a reallocation of existing spend rather than a new cycle of ad-market expansion, according to a September deep-dive from Morgan Stanley’s Asia internet team. The report, based on the sixth wave of the bank’s AlphaWise China Advertiser Survey, found that 72% of advertisers already using or planning to adopt AI said their total ad budgets remained unchanged, while 74% reported that cross-platform ad spending is becoming more concentrated.
The implication is stark: AI is not yet creating new advertising demand. It is changing how existing budgets are deployed, channeling spend toward platforms that offer higher conversion efficiency and more complete data feedback loops.
Efficiency Gains Arrive Before Revenue Expansion
Advertiser behavior underscores this early-stage dynamic. In 2025, 74% of surveyed advertisers used AI for content generation and 62% for campaign management — mature, efficiency-focused use cases. By contrast, penetration for product feed and catalog optimization stood at just 9%, and generative engine optimization at only 5%. These AI-native formats, which sit closer to discovery and purchase decisions, remain largely untapped.
Half of advertisers (50%) said AI has already improved return on investment, and 46% cited better handling of customer inquiries and pre- and post-sale support. But the budget-level data tells a more cautious story: AI is currently a tool for doing more with the same spend, not for unlocking new spend.
From Budget Migration to Transaction Closures
Morgan Stanley projects that by 2030, separately identifiable AI monetization will reach roughly RMB 291 billion, comprising about RMB 10 billion in candidate-set advertising and RMB 281 billion in AI-attributed transaction commissions. Critically, this figure excludes revenue gains from predictive AI improving the monetization efficiency of existing ad inventory — those gains would still show up as traditional advertising revenue.
The real expansion comes later. By 2040, as AI agents participate in discovery, decision-making and transaction execution, AI-facilitated advertising revenue could reach RMB 319 billion, with AI-attributed transaction commissions potentially hitting RMB 12.4 trillion. After accounting for cannibalization of traditional businesses, incentive spending and partner revenue sharing, however, the net market expansion will be smaller than these headline figures suggest.
The two phases reflect different commercial logics: before 2030, it is about budget migration; around 2040, it is about a genuine reconfiguration of business models and market size. The 2040 projection is best treated as a long-term scenario anchor rather than a precise timeline.
Who Captures the Value? The Four Control Points
As AI evolves from an ad-delivery tool into an agent capable of understanding needs and executing tasks, platform competition will center on control of the commercial chain. Morgan Stanley identifies four key nodes: who initiates demand, who determines the candidate set, who completes the transaction, and who owns the resulting data and write-back permissions. From advertising and sponsored ranking to transaction commissions, payments and merchant services, each link corresponds to a different value-capture model.
This also explains the difference between standalone AI and embedded AI. Standalone AI can aggregate cross-domain user intent, but embedded AI — closer to products, merchants, payments and transactions — is better positioned to convert user needs into quantifiable commercial outcomes. Survey data shows 47% of advertisers believe AI-created value will be shared among participants, while 37% see platforms as the largest single beneficiary.
Key Takeaways for Investors
- Near-term, AI is a share-shift story, not a market-expansion story. Platforms with superior conversion efficiency and data feedback — likely the largest incumbents — are the primary beneficiaries of budget consolidation.
- Advertiser AI adoption is still concentrated in content generation and campaign management. Watch for rising penetration in product feed optimization and generative engine optimization as leading indicators of AI-native ad formats taking hold.
- The 2030 monetization figure (~RMB 291 billion) is a migration estimate, not incremental revenue. Investors should separate AI-attributed revenue from efficiency gains that flow through traditional ad lines.
- Long-term value hinges on transaction-chain control. Platforms that embed AI agents into discovery, decision and transaction execution — capturing commissions and merchant services — could unlock revenue beyond advertising.
- AI may amplify, not erase, existing platform advantages. Differences in user demand, supply, transactions and data are likely to widen rather than narrow.
