From Free AI to AI That Makes Money... MOAI's Business Viability Put to the Test [MOAI: Expectations and Reality (5)]

[Financial News] Attention is focused on whether the homegrown artificial intelligence (AI) service MOAI can establish a sustainable revenue model. Since its basic services will be free, a key challenge is how to handle the operating costs that rise as usage increases. Participating companies plan to broaden their revenue streams through specialized services and business offerings for companies, but are still working on specific revenue models. Government regulatory reforms to support corporate data use and service integration are also seen as a factor that will determine commercial viability.■High expectations for free access, but turning it into revenue remains a challengeAccording to industry sources on the 28th, MOAI will launch a closed beta in October before rolling out its full service in December. The three consortia—SK Telecom, KT and Kakao—are preparing their businesses with a view to expanding their user base while keeping basic services free and securing separate revenue streams. 
The initial challenge is attracting users, but ensuring the service is profitable is essential to sustaining it. AI services use computing resources to process questions and generate answers, so as usage grows, the infrastructure and operating costs needed to handle it also increase. The more people use the service for free, the more important it is to establish ways to recoup those costs.
Users' willingness to pay is not particularly high. In a survey commissioned by Financial News from research and technology firm OpenSurvey, which polled 1,000 men and women aged 20 to 59 nationwide, just 24.7% said they would be willing to pay for MOAI's advanced features. Asked why they would want to use the service or were interested in it, 69.9% of respondents selected its being free. In other words, attracting interest by offering the service for free does not necessarily mean there is demand for paid features.
Participating companies are exploring ways to broaden revenue streams while reducing costs. For example, SK Telecom plans to select AI models suited to users' requests and use lightweight models to cut computing costs. It also plans to expand its revenue streams in the future through specialized services and partnerships. 
The key is how to connect free services with revenue-generating businesses. Participating companies will have business opportunities only if the service goes beyond simply providing AI answers to help with specialized tasks or lead users to other services. The costs to be borne and revenue to be earned by service operators, model developers and specialized service providers must also be determined. An industry official said, "More people using a free service and the business generating revenue are separate issues," adding, "We need to build a model that leads users to specialized services and decide how participating companies will share the costs and results."
■To expand into applications and payments, data and accountability standards must be clear Expanding services to broaden revenue streams also raises issues around data use and user protection. To move beyond explaining public benefits and help people actually apply for them, services need identity verification, the issuance and submission of documents, and connections to the systems of relevant agencies. In fields such as finance and healthcare that handle sensitive personal information, the scope of data use for each service must also be clearly defined.
Government support must also lead to an institutional framework that enables companies to launch and operate real-world services. The tasks AI can perform and the requirements for using data should be reviewed across fields such as finance, healthcare, education and shopping, while relevant ministries should jointly provide specific standards where the interpretation is unclear. If the scope of what is allowed remains uncertain, companies may struggle to determine whether they can commercialize a service even after developing it, which could make investment and service expansion more difficult.
Industry voices have called for new services to be tested in controlled environments through measures such as regulatory sandboxes, with the results used to improve the regulatory framework. The aim is to clarify the conditions under which companies can expand their businesses while assessing the services' effectiveness and risks.
User protection standards also need to be updated. Even with consent for data use and confirmation procedures in place before actions are taken, errors in bookings or payments and personal information leaks may still occur. An industry official said, "For AI to go beyond providing information and proceed to actual applications and processing, connections between agencies and companies and standards for data use must be in place," adding, "The government needs to reduce institutional uncertainty encountered on the ground and support services in becoming viable businesses."
[email protected] Yoon Hong-jip Reporter