Apple's AI Strategy Faces Setbacks, Siri Upgrades Encounter Challenges

2025-03-04

In recent years, the rapid advancement of generative artificial intelligence has driven up the market values of tech giants like Microsoft and Nvidia, even surpassing $3 trillion to become the most valuable companies globally. However, in this AI boom, Apple seems to have maintained a certain distance, adhering to the philosophy advocated by its CEO Tim Cook: "Not the first, but the best." Apple often stands out through meticulously designed and polished products rather than rushing to seize market opportunities.

This strategy is also reflected in Apple's AI approach. At the end of last year, Apple introduced an AI strategy named Apple Intelligence to compete with rivals such as OpenAI’s ChatGPT, Microsoft’s Copilot, and Google’s Gemini. Despite generating significant attention at launch and even temporarily boosting Apple’s market value above that of Microsoft and Nvidia, it soon revealed a series of issues.

According to insiders, Apple Intelligence might already be a failure before its widespread release. The AI strategy Apple unveiled in June 2024 included a range of exciting features like writing tools, voicemail transcription, Genmoji, and the Image Playground app. However, the rollout of these features was slow, being released in waves, making the entire plan seem lacking in urgency and appeal.

Meanwhile, OpenAI’s ChatGPT achieved remarkable success in the AI field, quickly winning over a large number of iOS and Android users. Although Apple partnered with OpenAI to integrate its technology into Siri to enhance its AI efforts, Apple Intelligence did not perform satisfactorily. Internal data shows that despite Apple’s claims of high iPhone sales in regions where AI features were available, the actual usage rates were at historical lows.

Apple once demonstrated immense potential in the AI domain, especially after launching the Siri digital assistant in 2011. However, in recent years, Apple appears to have fallen behind evolving trends and lags behind competitors' offerings. The release of Apple Intelligence could have been an opportunity for Apple to boost iPhone sales in the sluggish Chinese market, but the slow rollout and lackluster features only worsened the situation.

Moreover, Apple faced criticism for generating incorrect news alerts with its AI summary feature, forcing the company to temporarily disable the feature and implement a permanent fix. These issues further compounded Apple’s challenges in its AI strategy.

Despite these setbacks, Apple is still striving for redemption. According to Bloomberg reporter Mark Gurman, Apple plans to upgrade Siri through an upcoming software update. The update is expected to include a next-generation version of Siri, unlike its predecessor, featuring a unified core. However, due to time constraints and technical challenges, Apple failed to launch this version of Siri as planned. Gurman claims that Apple might roll out its AI strategy via the iOS 19 software update as initially planned, but the new system, internally called “LLM Siri,” will likely not be unveiled until June 2027.

Following the setbacks in Apple’s AI strategy, insiders raised concerns, noting that a truly modern, conversational version of Siri might not reach consumers until iOS 20 in 2027. This indicates that Apple is about five years behind its competitors in the AI field. Additionally, Apple faces key issues like the lack of advanced GPUs, as demand for Nvidia chips from tech giants like Microsoft and OpenAI continues to rise. Although Apple is addressing this issue by developing its own solutions through its chip team, the situation is exacerbated by competitors poaching Apple’s talent.

To bolster its AI efforts, Apple recently transferred veteran project manager Kim Vorrath to its Siri and AI department. According to a leaked memo from Apple’s AI head John Giannandrea, the AI department will focus on overhauling Siri’s infrastructure and improving the company’s internal AI models. Whether Apple can catch up in the AI domain and achieve self-redemption remains to be seen over time.