Businesses moving from AI experimentation to operational execution
With the theme "the New Brand Era", iBrandi Festival 2026, was held in Beijing on July 2 and 3.
The fourth edition of the festival featured a new artificial intelligence business forum, which attracted representatives from both domestic and overseas companies, including Google Cloud, Microsoft, DeepZero and ModelBest Inc.
In the forum, the participating entrepreneurs and industrial insiders discussed how AI can reshape corporate growth pathways, focusing on topics including AI-powered searches, agent workflows, organizational collaboration and edge intelligence.
Experts agreed that AI development has passed the tipping point of conceptual showmanship this year. Companies need to move beyond the conceptions that “AI equals cost reduction” or “AI equals personal efficiency improvement” and truly embed AI into their entire business chain to build differentiated competitiveness.
Huang Xiaonan, the founder and the CEO of DeepZero, pointed out three major bottlenecks in AI implementation: fragmented applications, misplaced value and lack of infrastructure.
“The key to breaking through lies in introducing agentic software that embeds multiple AI agents into sales, customer insights and marketing communications — shifting from completing isolated tasks to supporting complete business objectives and achieving ‘AI for Growth’,” he said.
Li Chenzhong, the AI product solutions director at Alibaba’s Tongyi Qianwen, noted that model capabilities are becoming increasingly homogenized. He advised companies not to obsess over model selection, as the dividing line has shifted from models to application scenarios. The real priority, he said, is to quickly validate business hypotheses and implement scenarios using mature ecosystems.
Significant changes are also taking place on the brand side. Zhang Qiang, the head of retail for Google Cloud Greater China, introduced the concept of “Agent as Brand”, suggesting that AI is transforming brands from static symbols into interactive, perceptible service carriers.
By continuously gathering user feedback data, brands can build long-term competitive advantages through AI-powered touchpoints.
Addressing brand exposure challenges in the AI search era, Li Jiao from Qihoo 360 proposed that generative engine optimization is not about competing for rankings, but about building brand AI awareness assets. By establishing knowledge bases, knowledge graphs and trusted source systems, brands can become more easily understood and recommended by large language models.
Organizational efficiency improvement also requires moving beyond the mindset of “replacing individual components”.
Gu Dawei, a global partner at Microsoft and the general manager of Microsoft Mobile China, noted that while AI has significantly boosted personal productivity, organizational gains remain limited because AI has not yet been properly integrated into existing workflows.
AI should become a tool for aligning team understanding rather than merely a personal productivity plugin, he said.
Liang Chao, the head of product and operations for Plaud China, emphasized the importance of converting high-frequency communications and business data into reusable context. Only then, he said, can AI evolve from a one-time tool into a sustained business partner.
Meanwhile, edge-side large language models developed by companies such as ModelBest are bringing AI from the cloud to physical devices such as smartphones, automobiles and robots. Virtual idols and other AI-native IPs are also exploring commercial pathways for personalized services, signaling that AI applications have fully transitioned from virtual scenarios into the real world.
The gathered industry insiders agreed that Chinese companies have strong advantages in application scenarios and data accumulation, which provide a solid foundation for AI adoption, and that the key to making AI truly useful for industries is to avoid flashy demonstrations and instead focus on real growth and solving specific problems.
