Companies involved in or considering participation in government artificial intelligence (AI) projects are expressing concerns that they are spending more time on administrative tasks than on research and development (R&D) due to a series of calls for proposals and evaluation schedules. Each new solicitation requires the formation of new consortia and the preparation of business plans, and even after selection, companies must undergo repeated evaluations, increasing their burden.
According to industry sources, the Ministry of Science and ICT's 'Everyone's AI' initiative is moving forward with a timeline that requires consortia to be formed and proposals submitted about three weeks after an information session.
An official from an AI company considering participation in the Everyone's AI project stated, "After the business briefing, we have to recruit participating organizations and adjust roles within just a few weeks. Since different companies are involved in each project, forming a new consortium takes a significant amount of time."
Another source pointed out that the lack of clear direction in project requirements adds to the burden on companies. "Although the project names differ, there are often no clear explanations of what differentiating factors are required. We find ourselves guessing the government's desired direction while reading the announcements multiple times to draft our proposals," the source said.
The administrative procedures necessary for preparing proposals are also considerable. Last year, the public AI transformation (AX) project required over ten types of documents, including business plans, personal data self-assessment forms, and various confirmations. The business plan must detail annual performance goals and technology development plans, which takes a significant amount of time to prepare.
Even after selection, the burden continues. Many projects involve a structure where subsequent support is determined through phased performance evaluations, requiring companies to simultaneously engage in technology development and evaluation preparation.
Uncertainty surrounding the projects is another concern for companies. The Everyone's AI initiative is still pending confirmation of next year's government budget. The Ministry of Science and ICT plans to select two to three consortia to operate beta services and then conduct a comprehensive evaluation of performance and usage rates to decide on further support. However, it has not yet been determined whether new participants will be selected if any consortia are eliminated.
A representative from the Ministry of Science and ICT explained, "With about four months remaining until the beta service, we plan to clarify the method of subsequent support while advancing the project."
While companies agree on the necessity of fostering an AI ecosystem through government support, they collectively emphasize the need for clearer distinctions in roles between projects and a more predictable support system.
In a situation where the number of companies capable of developing AI services and R&D personnel is limited, the simultaneous promotion of similar government projects forces the same companies and personnel to participate in multiple solicitations, leading to a dilution of their capabilities in consortium formation, proposal writing, and performance evaluation, which only increases administrative burdens.
However, there are also positive assessments of the government's solicitation method. An official from an AI company that has participated in government projects noted, "Many companies have been able to enhance their technology and secure commercialization opportunities thanks to government support. While it is meaningful in terms of expanding opportunities for private sector participation, if the differentiation and operational methods between projects are improved, we can expect positive effects."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.