[AI Pick] Ministry of Food and Drug Safety introduces AI in drug review… Complaint citation function to be released next month

The Ministry of Food and Drug Safety is introducing artificial intelligence (AI) technology in drug approval and review work.

This is to reduce the burden on drug reviewers by automating repetitive tasks and shorten the review period by increasing approval and review efficiency.

Among the pharmaceutical AI review assistance systems that the Ministry of Food and Drug Safety is planning to build, external functions that can be used by civil servants are expected to be released early next month.

According to the pharmaceutical and bio industry on the 9th, the Ministry of Food and Drug Safety is developing an AI screening assistance system for pharmaceuticals.

The Ministry of Food and Drug Safety established the ‘Food and Drug Artificial Intelligence Transition Promotion Team’ as a dedicated organization in February to introduce AI technology in drug approval and review work.

The AI ​​system construction project began this year, and a total of 22.3 billion won was decided to be invested over three years until 2028.

Through this project, the Ministry of Food and Drug Safety builds an AI model and system foundation suitable for supporting drug review work.

Specifically, we plan to develop a support tool to summarize and translate data submitted by drug developers and draft 55 types of review documents.

It was also decided to establish an industry support system that would allow companies to check for errors before submitting data.

We plan to gradually expand the review targets for this system from generic drugs this year to drugs for which data will be submitted next year, and new drugs in 2028.

Although the drug for which data is to be submitted is not a new drug, it is an item that requires safety and efficacy review.

If this AI system is actually introduced in the field, it is expected to contribute to shortening the drug review period.

The Ministry of Food and Drug Safety previously proposed a goal of reducing the medical product approval and review period, which took about 420 days, to about 240 days, the world's shortest.

It is known that some system construction is currently in the final stages.

The AI ​​model specialized for drug screening is based on LG AI Research Institute's ExaOne 4.5.

The Ministry of Food and Drug Safety is encouraging the reviewer function of the system to be used internally.

In addition, the external function for civil complaints that can verify application data is expected to be released in early December.