The "State of Business AI in 2025" report reveals that more than $30 billion has been poured into generative artificial intelligence. Yet, a staggering 95% of corporate pilot projects in this domain have stalled before reaching the production phase. The primary roadblocks include the improper use of tools, difficulties in integrating systems into existing workflows, and a lack of memory and adaptability. This situation underscores the existence of a "GenAI gap," where the success of these projects hinges on the practical and effective application of the tools.

While general AI tools have the potential to boost individual productivity, their impact on overall company profits remains relatively limited. On the other hand, customized platforms, despite being tailored to specific needs, exhibit even higher rates of pilot project failure. Large enterprises tend to launch numerous pilot projects but are often slow in rolling them out on a wider scale. In contrast, medium-sized companies can typically transition from the testing phase to full-scale rollout in about 90 days, with a substantial 70% of their budget earmarked for sales and marketing efforts.

Some observers have raised concerns about the transparency and objectivity of the report's findings. Looking ahead, the industry is expected to shift its focus towards "intelligent agent AI," with the goal of constructing an "intelligent agent network" that can facilitate large-scale and consistent applications.