AI Startup Motif Disputes Government's 'Own AI Foundation Model' Evaluation

Motif Technologies, an AI startup, has formally challenged the results of the South Korean government's second evaluation for its 'Own AI Foundation Model' (Dokpamo) project. The company lodged an appeal with the Ministry of Science and ICT, demanding a review of the assessment and disclosure of detailed scores for each participant.

The dispute centers on Motif's exclusion despite reportedly topping benchmark performance metrics. In the second evaluation, which concluded on November 18, three entities, LG AI Research, SK Telecom, and Upstage, advanced to the next stage. These selections were based on a combined score: 40 points for benchmark performance, 35 for expert evaluation, and 25 for user evaluation.

This is not the first controversy surrounding the Dokpamo project. The first assessment phase saw criticism regarding the originality of the AI models. Now, the second phase is facing scrutiny over the evaluation standards and the transparency of the scoring process. The situation has intensified calls for the government to provide clearer explanations of its criteria and selection rationale, especially as it pursues multiple large-scale AI support initiatives.

Motif has declared its intention to withdraw from the third phase of the Dokpamo project, irrespective of the outcome of its current appeal. This decision underscores the growing frustration among some participants regarding the perceived lack of clarity and fairness in the government's AI development funding processes.

“

We are officially appealing the Dokpamo second evaluation results and demanding detailed scores and a re-evaluation.