Founded in 1876, Dai Nippon Printing Co., Ltd. (DNP) is one of the world’s largest printing companies, employing over 37,000 people globally. With a portfolio spanning Smart Communication, Life & Healthcare, and Electronics, DNP is guided by its brand statement, “Creating future standards.” and a commitment to connect people and society while advancing sustainability.
As part of this commitment, DNP has long embraced emerging technologies. In April 2023, the company made a strategic decision to adopt AI across the organization. By May, DNP had built a secure environment for enterprise-wide use. In February 2025, the company launched ChatGPT Enterprise across ten core departments. Within three months, results included:
- 90% of use cases with ChatGPT Enterprise showed measurable results
- 100% weekly active usage rate
- 87% automation rate in time reduction
- 70% knowledge reuse rate (custom GPTs)
- 10x increase in processing volume
Accelerating adoption through strategic deployment
To fully realize the benefits of generative AI, DNP targeted ten departments with the highest potential impact. The company established clear benchmarks: each employee should use ChatGPT at least 100 times per week, and over 50% automation rate for task time reduction.
“We drove adoption by making usage visible,” says Hiroyuki Otake, General Manager of ICT Control Office, R&D and Engineering Management Division. “Each team experimented, shared learnings, and iterated. That momentum created a scalable impact.” As a result, individual improvements spread across teams through custom GPTs and shared use cases, forming core patterns now driving business transformation.
Cutting patent research time by 95%
In the departments where ChatGPT Enterprise was introduced, the greatest impact was seen in the ICT research and development division. Yohei Ishida, General Manager of P&I Innovation Research and Development Unit, Advanced Business Center, led his team to automate and improve patent research and filing strategies, replacing manual tasks.
His team built the following workflows using ChatGPT Enterprise:
- Patent research: automated search, summarization, and classification, cutting research time by 95% and expanding coverage 10x
- Application strategy: identified key differentiators between DNP’s technology and competitors’ patents, reducing rejection risk and minimizing revisions
- Competitive analysis: generated first-draft reports automatically, reducing preparation time by 80%
Ishida notes that the impact goes beyond efficiency to quality. “In the past, patent applications depended heavily on individual judgment, with standards varying by person and department. With ChatGPT Enterprise, we can now make objective decisions, which has improved both the volume and quality of our filings.” By elevating IP strategy, DNP is strengthening the foundations of product uniqueness and long-term competitiveness.
Building Python scripts with zero prior experience
DNP’s research division promoting production technology advances QCD (quality, cost, delivery) innovation to enhance the value of existing products and services, and pursues the development of new products and services. In areas that require advanced analytical and evaluation techniques, DNP has significantly reduced the time traditionally needed for tasks such as operating experimental equipment for material evaluation, conducting measurements, and performing analyses by leveraging ChatGPT Enterprise.
Key outcomes include:
- Structuring information from English patents and equipment principles in three days, down from several months
- Enabling employees with no Python experience to generate and run code through ChatGPT Enterprise
A particularly notable use case involved employees with no prior experience in Python, who were able to generate code and analyze data without any learning cost. Development work that would traditionally take more than a year was implemented within just a few days. By combining these capabilities with researchers’ expertise and knowledge, new insights were discovered, delivering significant impact across the entire division.
“Even those unfamiliar with IT or hesitant about it have used ChatGPT effectively and achieved significant results,” said Takamasa Yoshizawa, Director of Integrated Manufacturing Innovation Laboratory, Technology Development Center, and Michiko Ito from Evaluation and Analysis Research Institute. They highlighted data analysis with Python code as the most successful use case.
Enhancing IT compliance and cloud operations
DNP is modernizing IT governance with ChatGPT Enterprise. Masahiro Kobayashi, General Manager of System Infrastructure Development Division, ICT Center, Information Innovation Operations, highlighted improvements in tasks that were once manual and inconsistent:
- External security audit: cut audit comparison time from 30 minutes to 5 minutes; reduced cryptographic suite selection from 3 hours to 1 hour
- Cloud security: completed the initial check of ~100 CIS Benchmark noncompliance items in 10 minutes instead of two person-days
- Review support: shortened requirement reviews from 1 hour to 30 minutes by referencing design policies and past records
“The model excels at collecting relevant data and generating clear output,” says Kobayashi. “That allows our teams to focus on decision-making instead of document comparison.” He adds that AI will not replace human oversight: “Verification and final checks remain the responsibility of people.”
Preserving institutional knowledge through AI
One of DNP’s biggest challenges is knowledge loss. Expertise often lives in the minds of experienced employees, or buried in analog documents.
Under the leadership of Isaku Osawa, General Manager of Technology Development at the Advanced Business Center’s AI Business Development Unit, DNP is now using AI to address this issue head-on.
His team uses ChatGPT Enterprise to structure and digitize unstructured data from paper manuals to historical quality logs. Once ingested, these records become part of an internal knowledge base that anyone can access via custom GPTs. The time required to define the data architecture was cut by 90%. The team also doubled the number of technical papers they could review.
“Our goal is to turn generational knowledge into digital labor,” Osawa says. That shift not only offsets labor shortages but builds long-term capacity for innovation.
Building a foundation for AI-native business operations
“AI agents will blend seamlessly into various situations, allowing everyone to benefit from AI without even being conscious of it,” says Otake. He envisions a shift from human and AI collaboration to a foundation where parts of business run through AI to AI interaction. As robotics advances, this trend will accelerate, leading to a future where physical AI works in the real world.
Looking ahead, Otake emphasizes that knowledge preservation will be critical: “We must convert information created for people into information AI can understand and ensure that knowledge is preserved and shared. Our goal is to improve productivity as we prepare for a shrinking workforce.” The aim is to codify frontline know-how and quality records into structured data so that AI agents and future physical AI can learn and apply them, reducing reliance on individual expertise and turning it into an enduring competitive advantage.
Under its brand statement, “Creating future standards,” DNP seeks to expand strengths in printing and information technologies and transform into an AI-native company that generates new standards for society.
