Rifan Muazin
The global technology landscape is experiencing a profound transformation as leading firms accelerate their efforts to launch advanced smartphones powered by sophisticated artificial intelligence agents, capable of executing complex tasks from ordering meals to drafting communications solely through voice commands. This burgeoning field was prominently showcased at the World Artificial Intelligence Conference (WAIC) in Shanghai, signaling a pivotal shift in how users interact with their mobile devices, while simultaneously igniting fierce competition and raising critical questions about the future of the app ecosystem.
The Evolution of Mobile Intelligence: From Assistants to Agents
For years, smartphones have integrated basic AI functionalities, primarily through voice assistants like Apple’s Siri, Google Assistant, and Amazon’s Alexa. These assistants, while convenient for simple queries and commands, largely operate within predefined parameters and often require users to specify applications for task execution. The advent of "AI agents" signifies a leap forward. An AI agent is a more autonomous and proactive software entity designed to understand user intent, anticipate needs, and perform multi-step tasks across various applications and services without explicit, step-by-step instructions. This represents a paradigm shift from a command-and-control interface to a more intuitive, conversational, and anticipatory interaction model.
The concept of "agentic phones" envisages a device where the AI agent becomes the central operating system, acting as an intelligent intermediary between the user and the myriad of applications and online services. This promises a seamless, hyper-personalized user experience, where the phone proactively manages daily routines, optimizes tasks, and provides relevant information, thereby significantly enhancing productivity and convenience. However, this revolutionary potential also introduces significant challenges, particularly concerning data access, platform control, and the established business models of major app developers.
Shanghai’s WAIC 2026: A Global Showcase for Agentic Innovation
The World Artificial Intelligence Conference (WAIC), held in Shanghai on July 18, 2026, served as a critical platform for unveiling these next-generation mobile devices. WAIC has long been recognized as one of the most influential AI industry events globally, attracting researchers, developers, entrepreneurs, and policymakers to discuss the latest advancements and future trends in artificial intelligence. China, with its robust digital economy and massive smartphone market, is at the forefront of this AI integration, driven by significant investments in AI research and development. The conference provided an invaluable glimpse into the cutting-edge innovations emerging from the East, demonstrating the rapid pace at which agentic AI is moving from conceptualization to commercialization.
At least three prominent firms used the WAIC stage to present their "agentic phone" prototypes and commercial offerings, underscoring the intense competition to capture this nascent market. This concentrated display of innovation at a major global event indicates that the race to define the future of the smartphone is now in full swing, with Chinese tech giants vying for leadership alongside their Western counterparts.
Nubia’s NaviX Ultra: A Frontrunner in the Agentic Race
Among the most anticipated unveilings was Nubia’s NaviX Ultra, a smartphone powered by Doubao, the immensely popular AI chatbot tool developed by ByteDance, the visionary company behind TikTok. Nubia’s announcement, accompanied by promotional images of the sleek handsets, boldly declared, "A new era of AI agent smartphones begins." This declaration highlights the company’s confidence in Doubao’s capabilities and its ambition to lead the market.
The journey to the NaviX Ultra began earlier with a limited release of a prototype dubbed the "Doubao Phone" in December of the previous year, which quickly sold out, indicating strong consumer interest and a readiness for advanced AI features. This initial prototype demonstrated remarkable capabilities, executing simple voice commands to perform cross-app tasks such as ordering food and comparing shopping prices. The seamless integration and intuitive control promised by the Doubao Phone offered a tantalizing glimpse into a future of effortless digital interaction.
However, the rapid success of the Doubao Phone also exposed a significant hurdle: the resistance from established tech giants. Days after its release, major players in China’s digital ecosystem, including e-commerce behemoth Alibaba, social media and gaming giant Tencent, and online retailer JD.com, restricted the built-in AI assistant’s access to their respective platforms. This coordinated action effectively hobbled the prototype’s core functionality, as the AI agent’s ability to operate across various apps was severely curtailed. ByteDance, in response, was compelled to disable the powerful AI tool in certain circumstances, particularly when transactions or payments were involved, to comply with the platform restrictions.
This incident vividly illustrates the fundamental conflict at the heart of the agentic AI revolution. Traditional app ecosystems thrive on direct user engagement, which translates into data collection, advertising revenue, and platform loyalty. An AI agent that can seamlessly navigate and execute tasks across these platforms without direct user interaction threatens to disintermediate the app providers, potentially diverting user attention and valuable data. As Kiranjeet Kaur, associate research director at US market intelligence firm IDC, aptly noted, gaining broad access to apps owned by other companies is a critical "sticking point" for AI agent devices. Platforms are inherently reluctant to cede direct user contact, fearing they "lose control to another party."
For the NaviX Ultra, Nubia has reportedly adopted a strategy focused on "collaboration" rather than attempting to force access into existing apps. This suggests a more conciliatory approach, potentially involving API integrations and formal partnerships with app developers, which could mitigate some of the friction experienced with the Doubao Phone prototype. The success of this collaborative model will be crucial for the widespread adoption of Nubia’s agentic smartphone. As of the report, AFP had contacted Nubia for further comment on these strategic adjustments.
Honor’s "Robot Phone" and StepFun’s STEPX Neo: Diverse Approaches to Agentic AI
Beyond Nubia, other Chinese manufacturers are also innovating in the agentic AI space, albeit with distinct approaches. Honor, a prominent smartphone brand, showcased its "Robot Phone" at WAIC. This device distinguishes itself with a unique interactive camera that flips up on a small robotic arm, adding a physical dimension to its AI capabilities. Honor describes its device as "companion-centric," emphasizing its ability to interpret human gestures and even "bop to musical rhythms," alongside standard functions like taking selfies and steadying videos. The "Robot Phone" is slated for release later this year and will embed an AI agent leveraging several AI models, some of which have been co-developed with Alibaba, indicating strategic partnerships aimed at enhancing its intelligent functionalities. This physical embodiment of AI suggests a broader vision for mobile devices that extends beyond purely virtual assistants.
Shanghai-based AI startup StepFun also made a significant splash at WAIC with the unveiling of its "AI agent-native smartphone," the STEPX Neo. Unlike devices that integrate an AI agent into an existing smartphone architecture, the STEPX Neo is designed from the ground up with AI agency as its core operating principle. According to a sponsored article in state news agency Xinhua, StepFun’s chairman Yin Qi highlighted "deep partnerships" established with several major Chinese platforms, including mobile payment giant Alipay and ride-hailing leader Didi. These strategic collaborations are designed to enable the STEPX Neo to provide comprehensive, one-stop support for a wide range of services, including travel bookings, everyday purchases, local services, office productivity, and video editing. These partnerships are critical for the STEPX Neo to overcome the app access challenges faced by others, demonstrating a path toward integrated, multi-service AI agents through formal agreements rather than attempting unauthorized access.
The Global Race and Emerging Challenges
The push for agentic AI in smartphones is not confined to China. Globally, major tech players are heavily investing in integrating increasingly sophisticated AI features into their devices. Google, for instance, has been enhancing its Pixel phones with advanced AI capabilities, such as the ability to book appointments or summarize web pages, often leveraging its powerful Gemini AI models. These features, while not fully "agentic" in the proactive, cross-app sense, represent steps towards more intelligent and context-aware mobile experiences.
In a notable development outside China, US startup Brain Technologies launched its agentic "Natural AI Phone" in Japan in April, in partnership with mobile giant SoftBank Corp. A demonstration given to AFP in April showcased the phone’s potential: it successfully messaged a contact to apologize for being late based solely on an audio command. However, the demonstration also revealed the nascent nature of the technology, as the device "often failed to carry out requests." This mixed performance underscores Kiranjeet Kaur’s earlier assessment that "agenting is everyone’s dream, but we haven’t reached there yet," largely due to the still "patchy" performance of current AI agent tools. The "Natural AI Phone," which connects with a handful of apps including the popular social network LINE, highlights the ongoing efforts to refine and broaden the capabilities of these devices.
Implications and the Future of the Digital Economy
The advent of AI agent smartphones carries profound implications for the digital economy and user interaction. The core challenge remains the tension between the seamless, integrated experience promised by AI agents and the existing, fragmented app-centric ecosystem. If AI agents become the primary interface, mediating user interaction with various services, they could disrupt established business models that rely on direct app engagement, in-app advertising, and platform-specific data collection. Companies like Alibaba, Tencent, and JD.com are keenly aware of this threat, as losing direct user contact could significantly impact their market positions and revenue streams.
Data Privacy and Security Concerns: With an AI agent potentially having access to a user’s data across multiple applications—from communication logs to purchasing habits and location data—the implications for data privacy and security are immense. Robust security protocols, transparent data usage policies, and user control over data sharing will be paramount to building trust and ensuring ethical deployment of these powerful tools. Regulatory bodies worldwide will likely need to adapt quickly to address these new challenges, potentially introducing new guidelines for AI agent development and deployment.
Technological Hurdles: Beyond market and regulatory challenges, significant technological hurdles persist. Developing AI agents that can consistently understand complex, nuanced human language, accurately infer intent, and reliably execute multi-step tasks across diverse and evolving application interfaces requires continuous advancement in natural language processing, machine learning, and contextual awareness. The "patchy" performance observed in early prototypes underscores that while the vision is clear, the path to flawless execution is still under construction.
Redefining User Experience and Monetization: The shift away from traditional app usage, as predicted by Marc Einstein of Counterpoint Research, who stated, "in five or 10 years time, we won’t be using apps on our phones like we do today," necessitates a rethinking of user experience design. The focus will move from individual app interfaces to the overall agentic interaction. Furthermore, new monetization strategies will emerge. Instead of app downloads or in-app purchases as the primary revenue drivers, value might shift towards premium AI agent services, subscription models for enhanced agent capabilities, or partnerships based on facilitated transactions.
Conclusion: A Race with High Stakes
The competitive landscape for AI agent smartphones is currently fragmented, with "no clear winner in this race yet," as Marc Einstein highlighted. This intense competition is precisely why it remains "quite a hot topic" in the tech world. The potential to fundamentally change the digital economy and disrupt existing business models is immense. The demonstrations at WAIC 2026 clearly indicate that the future of mobile computing is inexorably linked to advanced AI agents. While challenges related to app access, technological maturity, data privacy, and regulatory frameworks remain, the relentless pursuit of seamless, intelligent mobile interaction by tech giants like ByteDance, Honor, StepFun, Google, and Brain Technologies signals a future where our smartphones are not just tools, but proactive, indispensable digital companions. The next few years will undoubtedly witness a fierce battle for market dominance, with profound implications for how we live, work, and interact with the digital world.
Sagoh
Lina Irawan
