The XGEN Agentic AI Platform has been awarded Grade 1 GS (Good Software) certification, Korea's national software quality certification. The certification review committee confirmed the highest grade on 6 July 2026.
What Grade 1 means
GS certification is a national software quality certification operated by the Ministry of Science and ICT under the Software Promotion Act. A designated accredited testing body tests and evaluates product quality — functional suitability, performance efficiency, reliability, security, and more — against the ISO/IEC 25000 family of international standards.
Grade 1 is the highest of those grades. For us the result means we did not claim performance; we had it verified by accredited third-party testing. AI software in particular gets asked questions that carry a different weight than traditional software does — the basis for an answer, reproducibility, safe behaviour when things fail. Those questions were answered against a national certification standard.
What changes for a company weighing adoption
GS certification, and Grade 1 in particular, works as concrete grounds for trust in public-sector and enterprise adoption.
- Eligibility to apply for Excellent Procurement Product designation with the Public Procurement Service
- Mandatory separate procurement when public software projects are put out to tender
- Designation for priority purchase by the Ministry of SMEs and Startups
- Technical certification points in public bidding — on many RFP evaluation sheets, GS Grade 1 or 2 counts as a valid software certification obtained within the last three years and earns bonus points (2 points, for example)
In other words, Grade 1 goes past the impression of "a good product" and becomes a verifiable qualification in the quantitative scoring of procurement and bidding.
Toward Enterprise AI that can be verified
The lab's principle has not changed — performance is proven by verification, not by assertion. Grade 1 is what that principle produced, and one knot tied in the "verifiable Enterprise AI" we are working toward.
Over a month or so of testing, everyone in the lab took part — document review, functional testing, regression testing, and explaining the product and handling errors on site. Physical certificates and certification documents will arrive in sequence, and the current status is reflected in Applied AI — certification and quality.