Tencent has finally placed a major bet on AI, but I do not know whether the capital market will recognize it.

Tencent has just released its 2026 Q2 financial report, and I will not go into details about the revenue and profit growth figures, as the announcement has documented them very clearly. After a quick read-through of the announcement, two figures left the deepest impression on me:

The capital expenditure this quarter hit as high as 52.784 billion yuan, up 176% year on year and 65% quarter on quarter. Is Tencent likely to be the Chinese internet giant with the highest level of capital expenditure at present? That depends on Alibaba's upcoming financial report.

The impact of new AI products on Non-GAAP operating profit is 10.5 billion yuan, which means that Tencent has added a net input of more than 10 billion yuan in a single quarter for products such as Hunyuan, Yuanbao, WorkBuddy, CodeBuddy and Xiaowei.

Tencent has finally made the decision that it should have made more than a year ago — to go all-in on AI. Of course, it is not too late to make this move now, as AI has not yet threatened Tencent's internet application ecosystem based on social media and games. The huge investment in the past quarter has produced at least one immediate effect: WorkBuddy is now the most popular AI office application in China, bar none. According to third-party estimates, the monthly active users of WorkBuddy on PC should be between 20 million and 30 million at present, although Tencent's management did not disclose the specific figures on the conference call.

The success of WorkBuddy has led to adjustments by competitors: Alibaba launched Qwen Office, ByteDance integrated Feishu into Doubao, and so on. This fully demonstrates what a huge and far-reaching impact WorkBuddy has had on the entire industry. However, it must be pointed out that this victory is still isolated for now. Tencent has made certain progress in the foundation large model, but such progress is far from consolidated; "Xiaowei" has been in gray release for more than a month so far, with an extremely low proportion of users being included in the test; Yuanbao cannot even rank among the top four C-end AI applications. WorkBuddy has changed the capital market narrative that "Tencent is not good at making AI products", but it has only made limited changes to Tencent's overall strategic position in the AI battlefield.

What are the reasons for WorkBuddy's success? I think there are three points. First, this product is very well-made, and Tencent's product manager culture has played its role. Second, it is closely connected with the WeChat ecosystem, which is probably the most important point. A friend of mine once said: "WorkBuddy plus the WeChat ecosystem is invincible, and it is difficult for anyone in China to defeat it." Third, Tencent has invested a huge amount of channel promotion budget, and I suspect that more than half of the more than 10 billion yuan incremental investment in the last quarter was put into WorkBuddy.

At the earnings call, Tencent's management emphasized: "This round of investment in AI-native business is mainly one-time, which will continue from the end of this year to the beginning of next year, but will not be maintained continuously in the future." In fact, if we do not take into account the incremental investment in new AI products, Tencent's Non-GAAP operating profit this quarter will grow by as high as 19% year on year, which is very rare in the current economic environment.

So here comes the question: Will the capital market appreciate Tencent's "double down" move on AI (the increase in capital expenditure and operating expenditure)? You know, since the second quarter of 2025, investors have been complaining that Tencent's investment in AI is insufficient, far behind Alibaba and ByteDance, which is the most important factor restricting Tencent's market performance, bar none. But in the past more than a month, the trend has changed again: from the United States to China, investors are paying more and more attention to the cash flow of tech giants, and are no longer interested in stories like raising capital expenditure, building data centers, and burning money on AI products. A year ago, tech giants splashing money on buying GPUs was a huge positive signal, the more they bought, the happier the capital market was; now the situation has become more complicated. Google's experience after releasing its financial report is the most obvious example.

Now Tencent's situation is a little similar to Google's: the rapidly rising capital expenditure has eaten up the operating cash flow, resulting in a negative free cash flow in a single quarter. This situation also occurred to Alibaba last year (of course, Alibaba was also affected by the instant retail war). It is well known that the capital market has always held the attitude of "wanting everything at the same time" for tech giants: wanting sound growth in the main business, huge investment in AI, and free cash flow for dividends and share buybacks at the same time. As long as the growth rate of capital expenditure exceeds the growth rate of operating cash flow, it is impossible to achieve "wanting everything at the same time".

If Tencent had done what it is doing now three to four quarters ago, the capital market would definitely have cheered, and perhaps its stock price would have hit a new high long ago (of course, in that case, it would also have encountered major setbacks in July this year). The current situation is more complicated, because the capital market is in a period of weak confidence and highly differentiated expectations. No one knows what investors value most at present, free cash flow or the growth of AI business.

In any case, I agree with Tencent's decision to increase its bet on the AI track. No matter how the capital market performs tomorrow, it should not change this decision, because this is the only correct decision at present. A great company should not succumb to the short-term pressure of the capital market, especially should not guess "what the capital market wants today, what it wants tomorrow, and what it wants this day next year". All companies that have made such guesses are no longer great, and I don't need to name them one by one, right?

This article has not received any funding or endorsement from Tencent or any of its competitors.

The author of this article does not hold Tencent's shares at present, but may hold them through funds or trust plans.

This article is from the WeChat Official Account "The Internet Phantom Thieves Group" (ID: TMTphantom), written by Pei Pei, Head of the Phantom Thieves Group, and authorized for release by 36Kr.