Mobile Agents Gain Momentum as Tech Giants Race for Innovation

Global tech giants Google and Samsung have recently launched the Galaxy S26 series flagship smartphone, hailed as a "benchmark model for intelligent agents" in the industry. Its standout feature is an intelligent interaction system that can independently identify user needs, connect multiple apps in the background, and automatically complete multi-step complex tasks such as food ordering, ride-hailing and shopping without manual app switching. The core functions of this model are highly consistent with the Doubao Mobile Assistant technology released by ByteDance’s Doubao team three months ago, particularly in three key capabilities: system-level wake-up, cross-app automatic operation and background task execution.

AI is making a crucial leap from laboratories to daily life, with intelligent agents as the core carrier, evolving from "being able to chat" to "being able to act". As global tech giants including Google, Apple and Samsung enter the field one after another, "mobile intelligent agents" are transforming from a concept into real products. The "mobile intelligent agent" track has gained great popularity and will become a key battlefield for AI application landing. The future competition in the mobile ecosystem will no longer be a contest of single hardware or software, but a comprehensive competition of system-level intelligence, full-link services and inclusive applications, which will reshape the global intelligent terminal development pattern.

The rise of this technological wave is no accident; it aligns with the national "Artificial Intelligence +" strategic deployment and is an inevitable requirement for promoting the high-quality development of strategic emerging industries. Relevant national strategies propose that by 2027, the penetration rate of new-generation intelligent terminals and intelligent agents will exceed 70%, and by 2030, the penetration rate will surpass 90%.

The emphasis on the popularization and development of intelligent agents lies in their important value as a specific practice of "integrated cluster development of strategic emerging industries". They are not only a typical carrier for AI technology to move from laboratories to daily life, but also a key driver for the transformation and upgrading of the AI industry. The development of intelligent agents can effectively promote the strategic transformation of the AI industry from single-point breakthroughs in technological research and development to large-scale landing of the entire industrial chain, integrating AI technology into all links of the real economy and social operation, and accelerating the in-depth adaptation of AI to manufacturing, service industry, social governance and other fields. The development quality of the intelligent agent track directly determines the efficiency and depth of AI empowering new quality productive forces, and is closely related to the overall process of high-quality development of China’s digital economy.

Faced with the accelerating technological development and the competitive pattern where global tech giants are scrambling for positions, it is crucial to seize this technological window, take proactive and targeted actions, encourage the accelerated scenario-based landing of independent and controllable new technologies and products, and build an intelligent agent industrial ecosystem with Chinese characteristics.

During the "14th Five-Year Plan" period, China has made historic breakthroughs in basic innovation and scenario application of AI large models, with a series of globally influential large model products and applications such as DeepSeek, Doubao, Qianwen and Kimi emerging. Forward-looking deployments and market-oriented explorations have also been carried out in smartphone applications and intelligent agent internet. China boasts the world’s largest smartphone market, the richest application scenarios and a complete terminal manufacturing industrial chain. These unique advantages have laid a solid foundation for the development of China’s intelligent agent industry. Giving full play to these advantages, accelerating the scenario-based landing of independent and controllable technologies, and building a safe and controllable industrial ecosystem are not only a realistic choice for China to grasp the initiative in development and achieve technological self-reliance and self-improvement, but also a strategic necessity for China to participate in global AI governance, grasp the right to speak on rules, and avoid falling into a passive follow-up situation in the new track.

As technology advances rapidly, governance must keep pace. While mobile intelligent agents break application barriers and improve service efficiency, they also bring a series of new challenges. To this end, we should adhere to the principle of "regulating while developing and adapting regulations to industry progress", and improve the regulatory system with the times. A tolerant and prudent regulatory attitude, seeking a dynamic balance between encouraging innovation vitality, maintaining market fairness and protecting user rights and interests, can escort the stable and long-term development of the industry.

For example, the code of conduct for AI assistants should be further clarified, with "user authorization" and "transparent operation" as the core principles. Before an intelligent agent executes a task, it should clearly inform the user of the operation content and permission scope. At the same time, relevant laws and regulations on data security and privacy protection should be improved. At present, when users first open Doubao Mobile Assistant, they need to agree to a privacy policy pop-up to clearly understand sensitive behaviors and permission usage scenarios; before operating mobile phone functions, Doubao Mobile Assistant will ask the user for permission in the dialog box, providing options of "allow once", "allow always" and "deny". In addition, waking up Doubao Mobile Assistant on the lock screen requires authentication through face recognition, voiceprint or password to prevent unauthorized access. Controversies in the industry are not a bad thing; what matters is that each controversy becomes a catalyst for the formation of industry consensus and rule-building, promoting the continuous standardization and upgrading of the industry.

Openness and collaboration are the vitality of the intelligent agent ecosystem and the only way to promote the high-quality development of the intelligent agent industry. Born with the attribute of breaking cross-scenario service barriers, intelligent agents will inevitably have a profound impact on the traditional single App ecosystem. Faced with this transformation, all parties in the industry should abandon closed thinking and vicious competition, move towards open integration, and promote collaborative innovation among upstream and downstream of the industrial chain and interconnection across platforms. On the one hand, we should promote cooperation among large model enterprises, terminal manufacturers and application developers, share technological achievements, jointly build standard systems, break technological and ecological barriers, and enable intelligent agents to better adapt to different terminals and application scenarios. On the other hand, we should encourage small and medium-sized enterprises to participate in the development of the intelligent agent industry, cultivate diversified market entities, and stimulate the vitality of industrial innovation.