Sumitomo Chemical, LINE Yahoo, dip share secrets of embedding generative AI company-wide

The body presents a panel discussion where leaders from Sumitomo Chemical, LINE Yahoo, and dip converge on a shared thesis: generative AI should not be treated as a universal tool bolted onto every process. Instead, they advocate for redesigning workflows around the technology's strengths. Sumitomo Chemical's representative frames this as a matter of identifying which parts of a task are better suited to AI versus human judgment, using a human-in-the-loop approach to retain control over outputs.

LINE Yahoo's CTO adds a structural perspective, arguing that merely introducing tools does not drive organizational change. The real work lies in connecting AI to internal systems and data, and nurturing so-called 'champion users' — employees who may not yet use the tools daily but can evangelize from within. dip's lessons, meanwhile, focus on psychology and discipline, warning against both overthinking and over-promising AI adoption targets.

The survey's finding that only 14.0% of companies report generative AI as 'gradually becoming established' suggests that broad adoption remains a challenge for most organizations. The approaches described by these three firms, ranging from phased security-conscious deployment to letting natural usage emerge without preset metrics, appear intended to counter the common failure modes implied by this low figure. The implication, grounded in the body, is that sustainable adoption depends less on technology enthusiasm and more on deliberate integration and human oversight.