AI News

Tongzhicheng Appraisal: AI review false-positive rate at 25%, exploring a human-machine collaboration path for real estate valuation

21ic Electronics NetworkUpdated Nov 16 18:34

AI summary

Tongzhicheng Appraisal Group disclosed that mainstream AI review systems in the real estate valuation industry currently have a false-positive rate as high as 25%, with OCR recognition error rates for property ownership certificates exceeding 6%, and that the "AI initial screening + manual review" model actually increases time spent by 14%. The institution proposed a five-step implementation framework for industry intelligentization, arguing that AI should serve only as an auxiliary tool, while valuation conclusions and professional responsibility remain with licensed appraisers. A 12-person delegation from the Ministry of Housing and Urban-Rural Development, the China Real Estate Appraisers Association, and the Shenzhen Housing and Construction Bureau has visited Tongzhicheng, affirming its transformation path of "technology empowerment, professionalism as the foundation, and compliance first."

Body · AI translation

SHENZHEN, China, Oct. 22, 2026 /PRNewswire/ -- The wave of artificial intelligence is sweeping across industries, profoundly changing their operating models and ecosystem structures. As an indispensable decision-making reference in fields such as financial risk control, state-owned asset disposal, urban renewal, and judicial appraisal, real estate valuation stands at the crossroads of digital transformation and looks forward to technological innovation. However, the practical pain points of implementing algorithms and the industry's strong compliance and ethical attributes mean that industry intelligentization cannot simply pursue wholesale technological replacement.

As a leading domestic real estate appraisal and consulting institution, Tongzhicheng Appraisal Group is testing AI applications in its business while confronting the industry's real pain points; at the same time, it has received recognition from supervisory bodies including the Ministry of Housing and Urban-Rural Development, the China Institute of Real Estate Appraisers and Agents, and the Shenzhen Housing and Construction Bureau during their research visits. It is also combining the consensus views of authoritative experts and professional institutions in the valuation industry at home and abroad, striving to explore for the valuation industry a path of intelligent transformation that is neither reckless nor lagging, values professionalism, and holds the bottom line.

At a themed salon on "The Application of AI Technology in Real Estate Valuation" hosted by the Shenzhen Real Estate Appraisers Association, Tang Hanzhou of Tongzhicheng's Quality Control Department, Zhang Anqing of its IT Department, and Yunshang Company shared their insights together. Rejecting the blind technological optimism in the industry that one-sidedly exaggerates the prospects of AI applications, they used frontline practical data as solid support to deeply analyze the real pain points of AI implementation and build a dual-track industry cognitive system in which "technological leaps and rational foundation-building" proceed in parallel.

Citing measured appraisal data, Tang Hanzhou pointed out a reality the industry cannot avoid: the false positive rate of mainstream AI review systems in the industry currently stands at as high as 25%, for reasons including a 1–1.5 month lag in policy updates and an identification rate of less than 60% for non-standard assets. When faced with complex ownership forms such as housing converted under housing reform and shared-ownership housing, AI still frequently "misses the mark." The OCR error rate for property ownership certificates exceeds 6%; for certificates from multiple versions since 1990 as well as blurry or smudged documents, key fields such as area, age, and nature of ownership still require repeated manual verification. The "AI preliminary screening + manual review" model actually increases time spent by 14%, with efficiency falling rather than rising.

He revealed a key reality: at the current stage, the application of AI in real estate valuation is on the whole still in an early phase where it is "usable but not good to use." Systematic efficiency gains have yet to be achieved, and the industry has yet to form a closed-loop, optimized intelligent operating system in areas such as data quality, policy synchronization, understanding of complex scenarios, and human-machine collaboration processes.

"It's not that AI is incapable, but that we are not yet ready." Tang Hanzhou's judgment injected a necessary dose of sobriety into this technological fever. The truth about AI applications at the current stage: the direction of the technical architecture is correct, but implementation remains mired in a compound bottleneck of fragmented data, dynamic rules, and blurred human-machine responsibilities. This is precisely the issue the industry must confront head-on when promoting deep application of technology.

Based on objective data, Zhang Anqing proposed a five-step systematic implementation framework for industry intelligentization, refusing to copy general industry digitalization templates and providing the industry with a transformation path that is implementable, replicable, and holds the bottom line:

Step one: self-diagnosis. Tailor an intelligentization roadmap to the institution's own business structure, data accumulation, technical reserves, and risk control system.

Step three: prioritize the application of assistive tools. Start with low-risk, highly adaptable links such as draft report generation, automatic compliance verification, and basic data collection to quickly achieve results, accumulate practical experience, and build team confidence.

Step five, and the most critical step: strictly observe ethical and compliance boundaries.

The views of the two Tongzhicheng experts align precisely with the authoritative standards of the global valuation industry. RICS's AI application standards explicitly emphasize that AI can provide auxiliary support for valuation work, but all valuation conclusions, professional responsibilities, and compliance risks must be borne entirely by certified appraisers; technology can never become an excuse for professional exemption.

Tongzhicheng hopes to promote a gradual industry-wide consensus: at present, AI can serve as an effective application tool, and technology is forcing the industry to iterate, driving appraisers to transition to higher-level roles from data movers and report writers to value judgers, decision-making consultants, and risk controllers. The future real estate appraisal industry will certainly be a comprehensive contest of professional depth + AI collaboration capability + compliance and risk control capability.

The practical exploration of a leading institution and the consensus judgment of industry experts have also received high recognition and affirmation from competent authorities. A 12-member delegation of leaders from the Ministry of Housing and Urban-Rural Development, the China Institute of Real Estate Appraisers and Agents (hereinafter referred to as "CIREA"), and the Shenzhen Housing and Construction Bureau visited Tongzhicheng Appraisal Group for inspection and guidance.

Jiang Wanrong, former Vice Minister of Housing and Urban-Rural Development and President of CIREA; Chai Qiang, Executive Vice President of CIREA; Zhou Tao, First-Level Inspector of the Real Estate Market Supervision Department of the Ministry of Housing and Urban-Rural Development; and Wang Bifeng, Deputy Director of the Shenzhen Housing and Construction Bureau, among other leaders, participated in the inspection and research activities. Wang Bo, Chairman of Tongzhicheng Appraisal Group; Chen Ping, President; Zhang Jie, Executive Vice President; and heads of various departments received them, and the two sides held in-depth exchanges on enterprise operations and industry development.

President Jiang Wanrong said that the real estate appraisal industry is in a critical window period of transformation and upgrading and must conform to the general trend of digital development and adhere to the main line of high-quality development. It should not only actively embrace AI technological innovation but also safeguard the industry's core foundation of independence, objectivity, and fairness. He encouraged Tongzhicheng to continue to play its exemplary and leading role as a top institution, continue exploring paths for deep integration of AI and valuation business, improve industry standards for intelligent implementation and risk control systems, and export replicable and promotable practical experience for the transformation and upgrading of the national valuation industry.

The research team spoke highly of Tongzhicheng's philosophy and practice of staying true to the industry's original aspiration, deepening its professional field, and rationally advancing intelligent transformation. The two sides held in-depth exchanges on core topics such as high-quality industry development, standardization system construction, real estate data governance, compliant application of AI technology, and industry risk control.

The on-site research and affirmation by the competent authorities carry the significance of an industry bellwether. Tongzhicheng's AI transformation path of "technology empowerment, professionalism as the foundation, compliance first" aligns with the industry's regulatory orientation and high-quality development requirements, providing an authoritative reference for the entire industry to break free from "technology anxiety" and "concept involution."

Human-machine collaboration is the new normal; only rational deepening can define the future

In the short term, AI cannot replace the core professional value of appraisers; its core value focuses on liberating repetitive manual labor, standardizing basic operating processes, and improving data processing efficiency. In the medium and long term, AI will become a standard tool for appraisers, reshaping the industry's operating models and talent standards.

The intelligentization of real estate valuation has never been a game of "AI replacing people," but a new ecosystem in which AI empowers people, professionalism fulfills people, and compliance protects people. Human-machine progress together is already an irreversible new normal for the industry. Only by grounding in professional foundations, consolidating the data base, strictly observing ethical boundaries, and making good use of technological empowerment can one stay true to the original aspiration amid the wave of industry change.

Source

21ic Electronics Network