Baidu: from search engine to full AI stack — the company’s CFO on strategy and robotaxis

Baidu has traveled from China's search leader to a company with a full AI stack — from silicon to autonomous taxis. Over nearly thirty years, the company has fundamentally reimagined its business. CFO Henry He explained in a Bloomberg interview what makes this transformation remarkable and where the company is heading.

From Search to Full Stack

Founded in the late 1990s as China's answer to Google, Baidu was long associated exclusively with internet search. Today, the picture is fundamentally different. The company manages an entire vertical chain in AI: developing its own processors, building cloud infrastructure, advancing its Ernie language model, and testing Apollo Go autonomous taxis on real city routes. According to Henry He, this architecture emerged not from theoretical considerations, but from practical necessity. US export sanctions on chips forced Chinese companies to invest in their own semiconductor production. For Baidu, this became both a threat and a catalyst. Today, the company positions itself as the only Chinese player controlling all layers of the AI stack — from hardware to interface.

Ernie and the Bet on Token Efficiency

The flagship Ernie model has gone through several generations of updates and, according to the company, in a number of tests outperforms foreign counterparts in understanding Chinese language and local context. According to the CFO, Baidu is betting not on model size, but on "maximizing token value" — that is, achieving maximum results with minimal computational costs. This is a concept increasingly discussed by Western companies as well: as competition grows, raw computational power gives way to efficiency. For Baidu with limited access to cutting-edge chips, this is a forced necessity — and, judging by He's words, a workable strategy.

Chinese-Style Safety

He touched on a sensitive subject: how Chinese tech companies approach AI safety and alignment. According to him, the approach is fundamentally different from the Western one — emphasis is placed on practical applicability and state compliance, rather than academic discussions about existential AGI risks. Silicon Valley talks a lot about disaster scenarios. Baidu, according to the CFO, looks at the problem differently: AI should serve concrete human and economic tasks. This doesn't mean safety issues are ignored — simply that answers are formulated in a different regulatory context.

Robotaxi: Global Race

Apollo Go is Baidu's flagship autonomous vehicle division. The service has already completed millions of rides in several Chinese cities without drivers behind the wheel. According to He, the company has accumulated a massive dataset of real-world driving on Chinese roads, giving it a competitive advantage over players just entering the market. At the global level, the competition looks like this:

• Waymo (Alphabet) leads in the US and expands into new cities
• Tesla is preparing to launch Cybercab in 2026
• Pony.ai and WeRide are the main rivals within China
• Apollo Go is betting on scaling and reducing ride costs

He acknowledges that the race will be long. But for a company with a full stack — from chips to cloud and software — launching new autonomous routes is simpler than for competitors dependent on third-party suppliers.

What It Means

Baidu is one of rare examples of a search giant successfully transitioning into a full-fledged AI player. The key question: how scalable is this model beyond China amid the world's technological fragmentation. According to the CFO, in the coming years the company will focus on deepening AI integration, rather than geographic expansion.