China's Tech Giants Are Paying People to Use AI

They're Literally Giving Away Cars

China's biggest tech companies just turned the Lunar New Year into the most expensive AI user acquisition campaign in history. ByteDance, Alibaba, Tencent, and Baidu collectively spent over 45 billion yuan (roughly $6.5 billion) handing out cash, luxury cars, robots, gadgets, and digital red envelopes to anyone willing to download and use their AI chatbots. The goal: get hundreds of millions of Chinese consumers hooked on AI before anyone else does.

ByteDance's Doubao gave away over 100,000 prizes during China's most watched TV event, the Spring Festival Gala on February 16. We're talking actual cars and cash packets as high as 8,888 yuan ($1,280). Alibaba's Qwen launched a "3 Billion Yuan New Year Freebies" plan that included bubble tea for 0.01 yuan a cup, and within nine hours, more than 10 million orders had been placed. By day six, that number had exploded to 120 million orders, with nearly half coming from rural areas. Tencent's Yuanbao distributed 1 billion yuan in straight cash, no strings attached, and hit the top of Apple's App Store free chart within 14 hours. Baidu kicked off its own 500 million yuan campaign for its Ernie chatbot.

This isn't normal marketing spend. This is an all out war for the future of how a billion Chinese consumers interact with technology.

Why Now

The timing is deliberate. Lunar New Year is when almost everyone in China is on their phone, either sending red envelopes, watching the gala, or chatting with family. It's the single biggest engagement window of the year, and tech companies have used it for years to push new products. What's different in 2026 is that the product being pushed isn't a payment app or a game. It's an AI assistant.

The strategic logic is simple and brutal. Whoever captures the most users for their AI chatbot during this window gains a massive first mover advantage. Unlike social media or e-commerce, where people routinely use multiple apps, AI assistants tend to be sticky. Once you get used to asking Doubao or Qwen for help with your daily tasks, switching to a competitor becomes inconvenient. The companies know this, which is why they're spending billions to be the first AI assistant that Chinese consumers build habits around.

Charlie Dai, principal analyst at Forrester, told CNBC that the companies are "in a high stakes race to capture users and build developer ecosystems before rivals lock in market dominance." The 2026 red envelope war isn't about Lunar New Year. It's about who owns the primary interface between humans and AI for the next decade.

The Four Way Battle

Each company is approaching the war differently, and the strategies reveal a lot about their broader AI ambitions.

ByteDance is leveraging its entertainment empire. Doubao is integrated with TikTok's Chinese version Douyin, and the Seedance 2.0 video generation model dropped the same week. ByteDance is betting that AI plus content creation equals an unbeatable combination. The company is positioning Doubao not just as a chatbot but as a creative tool for the hundreds of millions of people already on its platform.

Alibaba is going the commerce route. Qwen's subsidies are tied directly to shopping, with the AI helping users find deals, place orders, and access services. The 0.01 yuan bubble tea promotion was designed to show consumers that AI can save them money in ways they can feel immediately. Alibaba committed the largest war chest at 3 billion yuan because it has the most to lose: if another company's AI becomes the default shopping assistant, Alibaba's core business is threatened.

Tencent is playing the social angle. Yuanbao is being pushed through WeChat, which already has over a billion monthly active users. Tencent doesn't need to acquire new users so much as convert existing ones into AI users. Its 1 billion yuan cash incentive is the simplest pitch: use the AI, get paid.

Baidu was technically the first mover, launching its 500 million yuan campaign on January 26, weeks ahead of its rivals. Baidu has the longest history in AI among the four but the weakest consumer platform. Its Ernie chatbot is technically strong but has struggled to gain traction against competitors with built in distribution advantages.

Beijing Is Watching

Not everyone is happy about the spending spree. China's top market regulator, the State Administration for Market Regulation (SAMR), summoned all the major players, including Alibaba, ByteDance, Baidu, Tencent, JD.com, and Meituan, and told them to stop what it called "involutionary" competition.

"Involution," or neijuan, is the Chinese term for destructive, zero sum competition where everyone pours resources into a battle that ultimately leaves all participants worse off. Beijing has used this framing before to crack down on subsidy wars in electric vehicles and food delivery. The regulator told the companies to "jointly uphold a fair and competitive market environment to foster innovation and healthy development within the platform economy."

The warning is significant because it signals that Beijing, while supportive of AI development, is worried about a repeat of the ride hailing and food delivery subsidy wars of the 2010s, where companies burned through billions acquiring users who left as soon as the subsidies dried up. The government wants sustainable AI adoption, not a bubble.

What DeepSeek Started

The intensity of this competition traces back to one company: DeepSeek. When the Hangzhou startup released its R1 model in January 2025, it proved that competitive AI could be built at a fraction of the cost that Western companies were spending. DeepSeek's success turned the Chinese AI market upside down by showing that you didn't need $100 billion data centers to build useful models.

That triggered a price war. API costs for AI models in China collapsed by over 90% in 2025. What used to cost dollars per query now costs fractions of a cent. The commoditization of the model layer pushed companies to compete on distribution instead. If the models themselves are cheap and roughly equivalent, then winning depends on who can get the most users locked into their ecosystem.

The Lunar New Year campaign is the logical endpoint of that shift. Companies can't differentiate much on model quality anymore, so they're differentiating on user acquisition. It's the same playbook that ride hailing apps used a decade ago, and everyone remembers how that ended: one winner (Didi) and billions of dollars burned by everyone else.

The Bigger Picture for Global AI

What's happening in China has implications far beyond its borders. While American AI companies are focused on enterprise contracts and premium subscriptions, Chinese companies are making AI free or even paying people to use it. The result is a fundamentally different AI adoption curve.

India already has 100 million weekly ChatGPT users, as Sam Altman revealed at the India AI Summit last week. But China is potentially onboarding hundreds of millions of AI users in a matter of weeks through these subsidy campaigns. If that scale of adoption translates into better data, better fine tuning, and better real world applications, the competitive dynamics of the global AI race could shift significantly.

CNBC reported that China's tech shock "threatens the U.S. AI monopoly" and is "just getting started." Whether that's hype or reality depends on what happens after the red envelopes stop. If these users stick around and actually integrate AI into their daily routines, China will have accomplished something that no amount of venture capital funding in Silicon Valley has managed: mass consumer AI adoption measured in the hundreds of millions.

What Happens When the Money Stops

The crucial question is whether the users acquired through cash giveaways will keep using these AI assistants once the subsidies end. History suggests many won't. China's previous subsidy wars, from Didi vs. Uber China to Meituan vs. Eleme, showed that the majority of price sensitive users churned once normal pricing resumed.

But there's a counterargument: AI assistants might be different. Unlike a taxi ride, which is a transactional experience, an AI assistant that learns your preferences and becomes part of your workflow creates genuine switching costs. If Alibaba's Qwen helps you shop smarter, or ByteDance's Doubao helps you create better content, the value proposition might outlast the subsidy period.

The next few months will be the test. Watch the user retention numbers after the Lunar New Year buzz fades. Watch whether Beijing's involution warning translates into actual regulatory action. And watch whether the Chinese approach of subsidizing mass adoption ends up being a more effective strategy than the Western approach of selling expensive enterprise tools to corporations. The answer will tell us a lot about where global AI is headed.