Alibaba is intensifying its efforts in the artificial intelligence sector by updating its flagship chatbot, Qwen, in a direct push to compete with leading global AI tools such as OpenAI’s ChatGPT. This strategic move saw the revised Qwen app replace the older Tongyi version, becoming accessible on major app stores last Friday. In its official app-store descriptions, Alibaba positions Qwen as "the most powerful official AI assistant for its models," serving as the primary gateway for users to experience its latest Qwen AI model.

The company also has ambitious plans to integrate advanced agent-style features into Qwen, aiming to assist shoppers across its vast e-commerce platforms, including Taobao, according to reports from Bloomberg. Alibaba, however, did not respond to requests for comments regarding these developments. This initiative underscores Alibaba’s commitment to expanding the utility and reach of its AI capabilities, a drive that has been ongoing for the past two years amidst a global surge in AI tool development.

Alibaba has firmly established itself among the prominent AI developers in China, alongside emerging players such as DeepSeek and Moonshot AI. A key part of its strategy involves advocating for an open-source approach, making its advanced models readily available for other developers to utilize and adapt. This dual focus on internal development and external collaboration reflects a broader industry trend towards democratizing AI access while fostering innovation.

The strategic push into AI appears to be yielding significant financial returns for Alibaba. In the most recent June quarter, the sales generated from its AI products demonstrated remarkable growth, achieving triple-digit rates for the eighth consecutive quarter. Simultaneously, Alibaba has aggressively cut the operational costs associated with using Qwen3-Max, its largest model, by nearly half, as reported by the South China Morning Post.

The Qwen3-Max model, a trillion-parameter system initially launched in September, was introduced with some of the highest pricing tiers on Alibaba Cloud. The company has since substantially reduced its lowest API rates, decreasing them from US$0.861 to US$0.459 per million input tokens and from US$3.441 to US$1.836 per million output tokens. Furthermore, users engaged in batch processing tasks during off-peak hours are offered an additional 50 percent discount, incentivizing broader adoption and usage.

This aggressive pricing strategy is set against a backdrop of escalating competition within China’s AI model market. The Qwen3-Max model recently achieved first place in a cryptocurrency investment contest, outperforming leading models from both the US and China, highlighting its capabilities amidst this competitive landscape. Numerous start-ups, including Moonshot AI, Zhipu AI, and MiniMax, have recently unveiled new systems, vigorously promoting their enhanced performance and cost-effectiveness, contributing to several rounds of price reductions across the Chinese AI sector.

The competitive environment has seen earlier price battles among major model developers, now extending into new areas such as specialized coding tools. For instance, Volcano Engine, the cloud computing division of ByteDance, recently launched a new coding agent priced at just 9.90 yuan (US$1.30) for the first month, further intensifying the market’s focus on affordability and specialized applications.

Beyond price, companies are also exploring innovative methods to attract and engage customers. Moonshot AI, which receives backing from Alibaba, introduced a promotional offer last Tuesday allowing new users to experience its Kimi K2 Thinking model for as little as 0.99 yuan. This unique campaign encouraged users to negotiate their own discounts directly with the Kimi chatbot, leading to online discussions and sharing of "prompt injection" techniques, where users attempted to trick the system into believing they were Moonshot employees to secure better deals.

The unconventional marketing tactic, however, quickly led to unforeseen challenges. Within hours of the offer’s launch, Moonshot AI reported that its chatbot had begun "hallucinating," necessitating the immediate intervention of engineers to rectify the issues. This incident highlights the complexities and potential risks associated with highly interactive and open-ended AI customer engagement strategies.

Alibaba's rapid advancements in AI have also drawn attention internationally. "Silicon Valley doesn’t want to admit it, but the symptoms are obvious: we’re witnessing a full-blown Qwen panic," commented marketing specialist Tulsi Soni on social media.

Concurrently, Alibaba has faced the need to defend itself against serious allegations. A report by the Financial Times cited a White House memo claiming that Alibaba had allegedly provided China’s People’s Liberation Army (PLA) with specific technical support. This support reportedly included granting access to sensitive customer data, such as IP addresses, Wi-Fi details, payment information, and various AI services. The memo further alleged that Alibaba staff had shared information concerning "zero-day" security flaws.

While the Financial Times stated it could not independently verify the memo's claims, it indicated that the accusations underscore growing concerns in Washington regarding the potential risks associated with Chinese cloud and AI providers. Alibaba vehemently rejected these allegations, labeling them "completely false" and questioning the motivations behind the memo's leak. A company spokesperson characterized the report as a "malicious PR operation" designed to undermine a recent trade deal between the US and China. The Chinese Embassy in the US also criticized the FT’s reporting, calling the accusations "groundless" and a distortion of facts.