Invest1 distinct publisher3 min readUpdated
The National Police Agency is reassigning 79 officers into overseas-leak units and doubling the dedicated teams in Seoul and southern Gyeonggi. Insiders were the source in 60 of 68 cases through July.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
South Korea's National Police Agency is putting 79 more officers into units that investigate the overseas leakage of key national technologies, a list that runs from semiconductors and artificial intelligence to secondary batteries and defense industry technology [1]. The dedicated investigative units at the Seoul and Gyeonggi Nambu (southern Gyeonggi) agencies, where cases cluster, go from one team each to two, with teams assigned to specific technology fields [2].
The framing matters as much as the headcount. Police say the reallocation reflects a view that tech leak crimes have outgrown the category of harm to an individual company and now threaten national industrial competitiveness and economic security [4]. The Security Investigation Bureau of the National Office of Investigation said on the 20th that the 79 officers come from reassigning existing personnel [3], which makes this a priority shift inside a fixed force rather than a hiring round. Placement will follow recent case-handling results and regional investigative demand [5], concentrated in the Seoul metropolitan area, where large-company headquarters and R&D staff sit, and in defense, semiconductor and battery clusters [6].
The caseload supports the reweighting. Last year police uncovered 179 cases, eight of them involving national core technologies, and arrested 378 people [7], up from 123 cases and 267 people in 2024, increases of 45.5% and 41.5% [8]. Thirty-three of last year's cases involved leaks abroad [9], about 18% of the total [10]. Through July this year, 144 people were arrested across 68 cases, with the case count flat against the same period a year earlier [11]. Arrests per case have held at roughly two in both periods [12].
For operators, the composition is the useful part. Insiders were the source in 60 of the 68 cases this year, against eight involving outsiders [13] - about 88% internal [14]. Small and mid-sized firms and others accounted for 59 cases, versus nine at large companies [13], so the exposure is heaviest where security budgets are thinnest [14]. By sector through July, machinery led with nine cases, semiconductors and the automotive and railway sector had six each (two of the semiconductor cases involving leaks abroad), electrical and electronics four, and shipbuilding and information and communications three each [15]. Those named categories cover 31 of the 68 cases [16]. Where the destination country was confirmed, China was the most common, with four [17].
Two operational details raise the compliance stakes. First, police say these cases need expert judgment on whether leaked technology is identical to the original, whether it is genuinely non-public, and what its economic value is, and that the damage often never surfaces on its own [18]. That is effectively a documentation test: a company that cannot show what it classified as non-public, when, and why, is a company whose loss may be unprovable. Second, provincial agencies will designate industrial security liaison officers to advise companies, universities and research institutes [19], and police will strengthen a system to convert criminal intelligence picked up during damage reports and consultations straight into investigations [20]. An advisory conversation is also an intake channel.
Police also plan deeper technology assessment and information sharing with the Ministry of Trade, Industry and Energy, the Ministry of SMEs and Startups, the Korea Intellectual Property Office and the National Intelligence Service, plus more training by technology field [21]. Investigations already tend to run long because of vast digital evidence and the need for cooperation with domestic and overseas agencies [22].
Watch the second half. Case volume was flat through July while capacity doubles in the two busiest jurisdictions; if the count rises without a change in underlying behaviour, the driver is detection, not deterioration.
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
South Korea's National Police Agency will add 79 officers to units dedicated to investigating the overseas leakage of key national technologies, including semiconductors, artificial intelligence, secondary batteries and defense industry technology.
The dedicated investigative units at the Seoul and Gyeonggi Nambu (southern Gyeonggi) police agencies, where cases are concentrated, will expand from one investigative team each to two and will designate teams for specific technology fields.
The Security Investigation Bureau of the National Office of Investigation said on the 20th that it would reassign existing personnel to add 79 officers to units handling tech leak crimes.
Police said the move reflects a view that, as competition among nations to secure advanced technology intensifies, tech leak crimes have grown beyond damaging individual companies into offenses that threaten national industrial competitiveness and economic security.
Police plan to assign the additional officers based on factors such as recent case-handling results and investigative demand by region.
Police will reinforce dedicated staff in the Seoul metropolitan area, where large-company headquarters and research and development personnel are concentrated, and in regions where national strategic industries such as defense, semiconductors and secondary batteries are clustered.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Single-outlet relay of a specific official announcement
Every figure traces to one report of one police announcement: the staffing action, the team expansion and the caseload statistics are all internally specific and mutually consistent (arrests per case and derived shares check out arithmetically), which raises confidence in fidelity to the announcement. But there is no second publisher, no primary document, and no independent audit of the police-supplied counts, and the sector breakdown accounts for only 31 of 68 year-to-date cases.
Committed internal reallocation, outcomes not yet observable
This is a concrete organizational change rather than a proposal: officer counts, the doubling of two regional units and nationwide liaison-officer designations are all specified, and existing enforcement activity is quantified. Adoption is held down because the officers are reassignments rather than added capacity, the liaison and interagency measures are stated as plans, and the year-to-date case count was flat at 68, so no post-change effect is yet measurable.
Mildly overstated framing over an internal reallocation
The underlying reporting is sober and numeric, but the scale language runs ahead of the substance: 'doubling' applies to two regional agencies going from one team each to two, and the 79 officers are moved from other duties rather than added, with no disclosure of what those units give up. Statistics that justify the move are police-supplied and outcome-free, while the flat 68-case year-to-date figure sits against the headline growth trend.
Agency announcing and justifying its own capability expansion
The sole information channel is the police organization whose staffing is at stake: it supplies both the threat statistics and the framing that tech leaks endanger national competitiveness, and an official closes with a call to treat the staffing boost as a turning point for economic security investigation capability. That is a clear institutional interest in a rising-threat reading of the data, though the disclosed figures include an inconvenient flat year-to-date count.
Coherent official record, unverified beyond one outlet
Confidence is moderate: the facts are specific, arithmetically consistent and attributed to a named bureau on a stated date, which makes misreporting of the announcement unlikely. It is capped by single-publisher sourcing, self-interested statistics with no outcome data, an incomplete sector breakdown, and forward-looking elements (liaison officers, interagency sharing, training) whose execution cannot yet be checked.
invest
China's crude imports fell to a 2016 low, and the self-sufficiency bill looks cheaper1 distinct publisher
security
Shasta County's Peters hire is an access-control failure, not a voting-machine one1 distinct publisher
invest
Japan's fifth month of faster exports is real AI demand wearing a weak-yen coat1 distinct publisher
invest
China posts its fewest first-half marriages since 1980, and net household formation falls faster1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
en.sedaily.com
1 article · August 19, 2026