Summary
I would like to organize the use of generative AI in business into two main categories: (1) General common and (2) Industry-specific.
(1) General common – Divided into three phases: “Information gathering” ~ “Processing and Analysis” ~ “Output,” the “Information gathering” phase includes searches (sometimes displayed AI Overviews) and questions on generative AI platforms directly. Next, in the “Processing and Analysis” phase, there are applications such as idea generation (drafting, bouncing ideas off) and various data analyses. Finally, in the “Output” phase, generative AI can be used to create presentations, documents, minutes, and more.
(2) Industry-specific – Let me take manufacturing (automobiles), finance (insurance), and telecommunications as examples. In “Manufacturing (Automobile)”, investments in engineering work, software development, and in-vehicle AI assistants are active. In “Financial industry (Insurance)”, underwriting, insurance claims payments, and investment in call centers are becoming more common. In “Telecommunications industry”, the three main areas are contact centers, network operations, and internal knowledge search.
Main text
Hello everyone. This time, I will write about the recently popular use of generative AI in business, from the perspective of a problem solver. First, I would like to organize the usage of generative AI into two main categories: (1) General common and (2) Industry-specific.
1. How to Use Generative AI in Business: General common
First, as a “General common,” we will organize usage that generally applies to anyone involved in business, regardless of industry or job type. Let’s broadly divide the workflow into “Input” (Information gathering ~ Processing and Analysis) and “Output.”
During the “Input (Information gathering)” phase, many people continue to search (Google!) as before. I think there are many cases that AI Overviews are displayed here (on average, about 50%, and depending on search terms, about 80%). This also falls under the use of generative AI. Also, you might directly access generative AI platforms (ChatGPT, Gemini, Claude, etc.) to ask questions.
In the “Input (Processing and Analysis)” phase, for example, you might ask generative AI to draft ideas as part of idea generation, or conversely, you might bounce your own ideas off AI to work through them. Also, generative AI covers most of the so-called advanced data analysis. I think it will greatly help streamline the complex analysis tasks themselves.
In the “Output” phase, it is possible to have generative AI draft materials that you usually spend time creating over time, such as presentations (PowerPoint), documents (Word), and meeting minutes.
2. How to Use Generative AI in Business: Industry-specific
Next, I will organize how generative AI is used by industry (for example, Manufacturing (Automotive), Finance (Insurance), Telecommunications), and further by value chain (by the way, I also asked generative AI to draft this sortie).
2-1. Manufacturing (Automotive)
In “Manufacturing (Automobile)”, investments in engineering work, software development, and in-vehicle AI assistants are active.
2-2. Finance (Insurance)
In “Financial industry (Insurance)”, underwriting, insurance claims payments, and investment in call centers are becoming more common.
2-3. Telecommunication
In “Telecommunications industry”, the three main areas are contact centers, network operations, and internal knowledge search.
3. Points to note when using generative AI
Generative AI is extremely convenient, but there are three main points to be careful about when using it.
3-1. Always perform visual inspections by humans
Generative AI is very convenient because it can answer anything immediately, and its accuracy is improving rapidly, but at this stage (as of 2026), I believe that “human visual checks” of output results are still necessary. You might think that checking every time will reduce the effectiveness of business improvements, but since it would be faster than making it yourself, be sure to take the time to check it carefully.
To facilitate visual checks, you can make the process easier by adding “Please also include a link to the information source” to your prompt.
3-2. Be Careful of Hallucination
This is also related to “visual checks”, while generative AI gives plausible answers, if you look closely, it sometimes includes incorrect points (= hallucination). To avoid hallucination as much as possible, it’s also a good idea to clearly state in your prompt, “If there is no relevant information source, please reply ‘No information source,’ and if you don’t know, reply ‘I don’t know'”. Also, don’t miss a “visual inspection”.
3-3. Pay attention to security
If you mistakenly include confidential information when asking generative AI, the AI may learn from that information and include it in responses to other companies, leading to information leaks. To avoid this, it is necessary to set generative AI to “do not learn” mode. Each platform has its own setup method, so please check it out.
Recently, more companies (mainly large corporations) are developing their own AI platforms. In such cases, using your own platform might be recommended (at this stage, it seems to be a bit inferior to platforms that are well-known for their features and usability).
Also, as a fundamental point, it is important to be careful not to input any “gray area” information that might qualify as confidential.
4.Closing
This time, I have organized how generative AI is used in business. As you have seen, generative AI has a wide range of potential applications. Although it is still in the development phase, it is advancing rapidly, so I hope you, problem solvers will actively utilize it and apply it to solving daily problems.
That’s all for this time, and I would like to continue from the next time onwards. Thank you for reading until the end.