Alibaba, Baidu, Tencent Fall on Cheap AI Competition

The economic issue is margin compression. China’s large platforms are pouring money into cloud infrastructure, chips and model training, but low-cost open-source releases from domestic competitors make it harder to defend pricing power in AI services and limit the returns on that spending. If model access keeps becoming cheaper, the market has less reason to assume that AI will quickly lift revenue per user or expand cloud margins for the dominant players.

For investors, the problem is that the trade has become less about AI enthusiasm and more about monetization discipline. Alibaba’s RSI reading of 63.0 suggests the stock is no longer in the overheated zone, but the broader trend still looks fragile after a move that left it more than 20% below its January peak. Baidu’s momentum is weaker still, with bearish technical setup confirmed by a negative MACD and a 45.7 RSI, while Tencent has failed to regain its longer-term trend despite a summer bounce.

The pressure also extends beyond individual names. A weaker yuan backdrop and a choppy risk mood in Chinese assets can amplify moves in heavyweight internet stocks that are closely tied to domestic growth expectations and overseas investor flows. With China growth sentiment sliding sharply in the latest Adalytica snapshot and U.S. megacap AI spending still under scrutiny, traders are increasingly treating cheap open-source releases as a direct threat to valuation multiples across the sector.

The next catalyst is whether Alibaba, Baidu and Tencent can show that open-source competition is driving usage rather than destroying pricing power. Until earnings or product launches prove that AI can support higher revenue, every new low-cost model release is likely to be read as another reason to sell the stocks.