Due to concerns about information disclosure and other issues, Japanese companies such as Softbank and Hitachi have begun to restrict the use of interactive AI services such as ChatGPT in business operations.
Softbank warned employees last month about using ChatGPT and other business applications: “Do not enter company identity information or confidential data.” Although Softbank has issued guidelines for using cloud services such as interactive artificial intelligence, these rules are still repeatedly emphasized in view of the increasing attention paid to chat robots recently. Softbank plans to formulate rules to specify which businesses can use the technology and which applications can be used.
On March 1 this year, ChatGPT developer Open AI released an application program interface that allows enterprises to pay for the integration of ChatGPT into their services. Open AI said that the data collected from ChatGPT will not be used to improve performance, but its website showed that when this service is used for free, “we may use the data provided by you to improve our model. In this case, the information entered, such as the product being developed, can be used to answer the questions of other companies”.
In addition to Softbank, Hitachi will consider formulating new ethical rules to regulate the use of interactive AI and whether data is used correctly. The company plans to improve the AI usage guide developed in 2021.
Last month, Fujitsu informed employees of legal and ethical issues related to ChatGPT and other interactive AI services, including information protection. The company has set up a department to decide the use of AI. In February 2022, Fujitsu established a governance office to deal with ethical issues in product development and commercial use involving artificial intelligence. The company will remind the public of the risks of interactive AI through this office.
In January this year, ChatGPT had more than 100 million users only two months after its release. Microsoft integrated ChatGPT in Bing Search last month, pushing the daily users of the search engine to exceed 100 million for the first time in a month. Moreover, the application of this technology in enterprise operation is expected to increase significantly. However, it is also worrying that more and more people and enterprises are using interactive AI, because this technology may spread false information and cause data leakage, leading to some companies prohibiting the use of this technology.
Mizuho Financial Group, Mitsubishi UFJ Financial Group and Sumitomo Mitsui Bank have banned the use of ChatGPT and other services in the company’s operations. Mizuho’s restrictive measures will prevent employees from accessing the website from work terminals, with the purpose of “preventing the disclosure of important information such as customers and financial transactions due to improper use of employees”. The IT company NEC also prohibits employees from using ChatGPT.
Atsuhiro Goto, professor of research at the University of the Japanese Academy of Intelligence and Security, said: “Interactive AI can be easily used on personal computers, which makes employees less cautious. There is a risk that confidential data is accidentally entered and used outside the company. These services should be tested by a special internal team to clarify the use rules and information that can be entered.”
Panasonic Connect, a Panasonic holding company, began to use interactive AI to create files after taking measures to prevent data leakage and other problems. In order to reduce the risk of data leakage, the company reached an agreement with Microsoft, the co-developer of the software, not to use the content for secondary purposes such as artificial intelligence training. At the same time, employees were also warned not to enter personal information.