Tezign Technology releases a collaborative plan for agent transformation, proposing to build enterprises' "sovereign intelligence"

Economist Robert Solow once observed a phenomenon: despite massive investment in information technology, macroeconomic productivity statistics in the 1970s and 1980s did not show a corresponding significant increase. Solow, a Nobel laureate in economics, gave this finding the name the "Solow Paradox." At the themed forum of WTCC 2026 West Bund International Technology and Consumer Carnival, which opened yesterday, Fan Ling, founder and CEO of Tezign and a professor at Tongji University, mentioned this paradox. In his view, value can only truly be released when companies redesign their workflows and organizational structures around new technologies. The themed forum, under the theme "In the AI Era, Reunderstanding People," was initiated by Tezign and co-hosted by MoSu Space. WTCC 2026 West Bund International Technology and Consumer Carnival has been held for two consecutive editions and has become an AI+ consumer technology event based in Xuhui, representing Shanghai and facing the world. Wei Lan, deputy district mayor of the Xuhui District People's Government of Shanghai, shared Xuhui District's practices and outlook in AI industry clustering and innovation ecosystem development, emphasizing that moving technological innovation toward industrial application cannot be separated from the support of an innovation ecosystem, nor from the pull of enterprises' real needs. Fan Ling, in turn, drew an analogy with the historical transformation from steam engines to electric motors, using the fact that individual employees use AI to boost personal efficiency while overall corporate performance has not grown in tandem. "Replacing equipment takes very little time, but it took 30 years for factories to reconstruct their entire production processes according to the logic of electric motors," he said. Corresponding to the present, it is easy for enterprises to rapidly deploy AI tools, but completing the reconstruction of the entire organizational system is difficult. The industry is in the stage of the systemic reconstruction chasm of AI-native transformation; AI transformation is not just a technical issue, but also an organizational one. As an industry unicorn founded 10 years ago, redesigning an organizational structure suited to the AI era is tantamount to "breaking bones," yet Tezign actually did so. It is understood that Tezign has established a Pod+Community dual-track system within its organization: Pod, as the solid line, is a delivery unit with high cohesion and low coupling, responsible for business results; Community, as the dotted line, is responsible for accumulating cross-Pod professional capabilities and context, helping everyone continue to grow. The two tracks operate in parallel, giving the organization both execution capability and growth potential. In fact, what Fan Ling is exploring is precisely a benign and sustainable new paradigm that can turn business transformation into agents. For enterprises, Fan Ling said they should build two world models. One is a subjective world model, understanding customers' preferences, choice logic and behavioral contradictions, and understanding the subjective world of "external people"; the other is an enterprise world model, used to accumulate the company's internal processes, resource constraints and decision consequences, and understand the "internal self." Based on the two world models, enterprises need to build long-horizon agents that can continuously enter workflows, enter systems, enter evaluation and enter long-term operation. He also said that enterprises should not rely solely on third-party model services, but should establish an autonomous and controllable intelligent capability system, moving from "renting intelligence" to building "sovereign intelligence." In his view, rented intelligence solves point efficiency problems, but enterprises' data assets, business context and judgment experience cannot be accumulated along with it, and these are precisely the most difficult-to-replicate and most core sources of enterprise competitiveness. Building sovereign intelligence means enterprises possess intelligent assets that can be accumulated, reused and evolved—models replaceable, data controllable, judgments explainable—so that intelligence continuously appreciates with every real action and truly forms its own intelligence compound interest. At the forum, the Agent Transformation Co-construction Plan was officially released. The plan is initiated by Tezign and open to universities, research institutions, enterprises, model and technology partners, and professional practitioners. It aims to carry out long-term research and practical validation around real problems in enterprise agent transformation. The significance of agent transformation is not to let enterprises connect to more intelligence, but to let enterprises form their own intelligence: possessing their own judgment systems, their own context systems, and their own intelligence compound interest. This is exactly what Tezign is doing. The plan unfolds around three core directions: understanding, judgment, action and evolution, exploring how agents can more deeply understand real people, professional experience and enterprise contexts; how they can participate in complex judgments without standard answers and establish reliable human-machine collaboration and evaluation methods; and how they can continuously advance real work and keep improving based on action results. In terms of co-construction methods, joint research will be carried out around real industry problems, project experiments and validation will be conducted in actual business, evaluation standards, methodologies and case collections will be jointly accumulated, and results will be output through research articles, industry reports, teaching cases and other forms.