KOSUNI KOREA
Korean News in English
Technology

Understanding the Latest Updates on South Korea's 'Dokpamo' AI Model Scoring System

Published September 19, 2026 · 3 views
Understanding the Latest Updates on South Korea's 'Dokpamo' AI Model Scoring System

Ministry Releases Detailed Scores for Dokpamo Second Evaluation Round

The Ministry of Science and ICT has published the detailed scores for the second round of evaluations under its Independent AI Foundation Model (Dokpamo) program. SK Telecom secured first place with a total score of 70.6 points, though Motif led in benchmark comparisons.

Following Motif's appeal regarding scoring discrepancies, the Ministry publicly released the detailed scores for Dokpamo's second evaluation round, confirming SK Telecom as the overall leader. Motif (65.8 points) ranked first in benchmark comparison but fell to fourth place in overall score, prompting their appeal.

Debate continues over Dokpamo evaluation results. Discrepancies between expert and user evaluations have emerged, with the government promising improvements to the assessment methodology.

SK Telecom's Top Ranking and Motif's Appeal

The Ministry released detailed scores for Dokpamo's second evaluation round on August 27, 2026. SK Telecom achieved a total score of 70.6 points, surpassing Motif (65.8 points) to claim first place. However, Motif led in benchmark comparison results, prompting the company to request detailed score disclosure and reconsideration.

Multiple media outlets reported that SK Telecom secured first place with a total score of 70.6 points in Dokpamo's second evaluation round, while Motif ranked first in benchmark comparison but was eliminated, leading to their appeal.

The government released the full scores following Motif's appeal, confirming SK Telecom as overall leader. However, Motif's top ranking in benchmark comparisons has kept debate over assessment fairness ongoing.

독파모 점수 공개된 내용을 정리한 최신 가이드 설명 이미지

Dokpamo Evaluation Results and Related Appeals

In Dokpamo's second evaluation round, SK Telecom, LG, and Ustage advanced to the final three-way competition, while Naver Cloud was selected in the security AI category. Motif, having secured first place in benchmark comparison but being eliminated, filed an appeal which the government rejected.

The Ministry of Science and ICT and KISA have published an AI security threat response manual, and also released the E2E Data Guidelines for autonomous driving AI training data standards. Dokpamo public evaluations used a mix of common questions and free experience formats.

Dokpamo's first evaluation results were announced on January 15, 2026. The AI security threat response manual was published on July 8, 2026, while the E2E Data Guidelines for autonomous driving AI training data standards were released on June 19, 2026.

Continued Debate Over Dokpamo Evaluation Methodology

Debate continues over Dokpamo's evaluation methodology. The discrepancy between overall scores and benchmark comparison rankings has raised questions about fairness in the assessment process.

Experts noted that the metrics used in Dokpamo evaluations were unclear, making it difficult to understand. Disagreements also emerged between expert and user evaluations. While the government has promised improvements to the evaluation method, specific details have not yet been disclosed.

The National AI Strategy Council announced plans to improve Dokpamo's evaluation methodology as a measure to address ongoing controversy over results. However, with unclear public scoring criteria and lack of detailed explanation, questions remain.

Recent Developments Regarding Dokpamo

On August 27, 2026, the Ministry of Science and ICT released detailed scores for Dokpamo's second evaluation round, confirming SK Telecom as overall leader. Motif, having ranked first in benchmark comparison but being eliminated, filed an appeal which was rejected.

Motif's appeal response was expected to be announced on September 10, 2026, though the government had already released scores and rejected the appeal by August 27. In Dokpamo's second evaluation round, LG, SK Telecom, and Ustage advanced to the final three-way competition, while Naver Cloud was selected in the security AI category.

The Ministry of Science and ICT and KISA published an AI security threat response manual on July 8, 2026. The E2E Data Guidelines for autonomous driving AI training data standards were released on June 19, 2026. Dokpamo public evaluations used a mixed format of common questions and free experience.

Key Information to Verify

Detailed scores related to Dokpamo evaluation results were published by the Ministry of Science and ICT, though specific explanations regarding the assessment methodology remain limited. The discrepancy between overall scores and benchmark comparison rankings suggests an issue with clarity in the scoring system.

Motif's appeal prompted the government to release scores, but no concrete plans for improving the evaluation method have been announced yet. Dokpamo's first-stage evaluation results were published on January 15, 2026, while relevant policies and manuals were released between June and July 2026.

The government's promise to improve Dokpamo's assessment methodology appears aimed at resolving ongoing controversy over the results. However, without clear explanations of scoring methods or criteria, questions persist regarding fairness and transparency.

Frequently Asked Questions

Why did a dispute arise in Dokpamo's second evaluation round where SK Telecom ranked first overall but Motif topped benchmark comparisons?

The discrepancy between overall scores and benchmark comparison rankings has raised questions about assessment fairness. Motif, having ranked first in benchmark comparison but falling to fourth place in overall score, filed an appeal which was rejected by the government.

Why does debate over Dokpamo's evaluation methodology continue?

The differing rankings between overall scores and benchmark comparisons, combined with unclear assessment metrics, have made it difficult for stakeholders to understand. Disagreements also exist between expert and user evaluations, further fueling the controversy.

What has the government promised regarding improvements to Dokpamo's evaluation methodology?

The National AI Strategy Council announced plans to improve Dokpamo's evaluation methodology but has not yet disclosed specific details. The government views this as a measure to address ongoing disputes over results.

Frequently Asked Questions

Disputes around the Ministry of Science and ICT's Dokpamo second evaluation results stem from unclear assessment criteria, with SK Telecom leading overall while Motif topped benchmark comparisons. The government rejected Motif's appeal but has promised to improve scoring methodology without providing specific details.

Dokpamo-related policies and manuals were sequentially released between January and July 2026, though debate over evaluation results continues. To resolve these disputes, the government must enhance transparency in its assessment framework to ensure fairness for both experts and users.

While detailed information on Dokpamo's evaluation results has been made public by the Ministry of Science and ICT, lack of clear explanation regarding scoring methodology leaves room for continued questions. The government needs to establish a transparent, understandable evaluation system that supports development in South Korea's AI sector.

Read More Related Articles

You can view more articles on similar topics here.

View Related Articles

FAQ

Why did a dispute arise in Dokpamo's second evaluation round where SK Telecom ranked first overall but Motif topped benchmark comparisons?

The discrepancy between overall scores and benchmark comparison rankings has raised questions about assessment fairness. Motif, having ranked first in benchmark comparison but falling to fourth place in overall score, filed an appeal which was rejected by the government.

Why does debate over Dokpamo's evaluation methodology continue?

The differing rankings between overall scores and benchmark comparisons, combined with unclear assessment metrics, have made it difficult for stakeholders to understand. Disagreements also exist between expert and user evaluations, further fueling the controversy.

What has the government promised regarding improvements to Dokpamo's evaluation methodology?

The National AI Strategy Council announced plans to improve Dokpamo's evaluation methodology but has not yet disclosed specific details. The government views this as a measure to address ongoing disputes over results.

Korean Source

This article is an English localization of a Korean-language report from BlogScore. Original: 독파모 점수 공개된 내용을 정리한 최신 가이드