Data from Rep. Cho Ji-yeon's office shows AI patent processing time near 2 years

Average review time hit 22.4 months in 2024 before easing slightly last year

Applications have risen by about 3,000 over four years as examiner count flatlines

Staffing falls short of authorized headcount, with only 10–15% holding specialist experience

[Getty Images Bank]
[Getty Images Bank]

AI-related patent applications are rising rapidly alongside the growth of the AI industry, yet the number of examiners handling them has hovered around 40 for years. The staffing shortfall has pushed the average processing time for AI-related patents to nearly two years.

The government has belatedly moved to expand its examiner workforce in advanced technology fields, but experts say it must build specialized review capacity at a pace that matches the AI industry's growth.

According to data submitted to the office of Rep. Cho Ji-yeon of the People Power Party — a member of the National Assembly's Trade, Industry, Energy, SMEs and Startups Committee — the average processing time for AI-related patents last year was 21.3 months.

The figure stood at 12.2 months in 2021, climbed to 16 months in 2023, and peaked at 22.4 months in 2024. It edged down to 21.3 months last year, but applicants still wait an average of nearly two years for a decision.

A growing backlog compounds the problem. The number of AI-related patents awaiting examination surpassed 10,000 for the first time in 2023, reaching 13,192, up from 7,850 in 2021. The figure rose further to 14,639 in 2024 before falling to 10,979 last year — still above the 10,000 mark for the third consecutive year.

Filings and examination requests are also climbing. AI-related patent applications rose from 6,250 in 2021 to 9,223 last year, an increase of roughly 3,000 over four years. The number of examination requests for AI-related patents reached 6,804 last year, up nearly 1,000 from 5,852 the year before.

The examiner headcount, by contrast, has barely moved. The number of AI patent examiners — including both permanent and fixed-term staff — grew from 35 in 2021 to 37 in 2022, dipped to 36 in 2023, then rose to 40 in 2024. It remained at 40 last year and, as of August this year, had slipped to 39.

Authorized positions are also going unfilled. Actual headcount fell short of the authorized ceiling in every year except 2024, when both figures stood at 40. This year, the authorized ceiling is 43 but only 39 positions are filled, leaving a staffing gap of roughly 10 percent. The share of examiners with relevant specialist experience has also remained low, holding at between 10 and 15 percent in recent years.

The mismatch between surging demand and stagnant staffing has drawn warnings that delays in patent examination could hamper both technology development and intellectual property protection.

"AI-related patent applications are increasing with the rapid growth of the AI industry, but the number of qualified examiners available to review them is simply not sufficient," said an official from the opposition camp.

The Korea Intellectual Property Office has also begun addressing the shortage. On Tuesday, it announced plans to hire 41 specialists through a competitive recruitment process aimed at raising the quality of patent, utility model, trademark and design examinations and cutting wait times. Of those, 27 experts in advanced fields — including AI, biotech and secondary batteries — will be recruited as Grade 6 examiners in the patent and utility model division.


mp1256@heraldcorp.com