Policy Brief: Formal Compliance, Hollow Substance

Why Korean Law Makes Genuine Horizon Europe GEP Compliance Structurally Impossible

Gender Watchdog — April 2026

Addressed to: European Parliament Committee on Industry, Research and Energy (ITRE); European Parliament Committee on Women's Rights and Gender Equality (FEMM); European Commission DG RTD; EIGE

The Finding

Korea became the first Asian country to associate with Horizon Europe in January 2025. Gender Equality Plans have been a mandatory eligibility criterion for all Horizon Europe participants — including Associated Country institutions — since 2022. As of March 2026, eleven of the twelve Korean institutions whose researchers participated in the Korea-EU Horizon Europe Advanced Biotechnology Researchers Networking Forum (June 2025) had no publicly accessible GEP.1 The twelfth filed one in March 2026 — its task force appointed in December 2025, the same month its 34 falsified international partnerships became a multi-language public record.2

This is not an administrative lag. It is the output of a legal architecture.

The Hypothesis

Two provisions of the Korean Criminal Act make genuine GEP compliance structurally impossible — not for individual institutions, but for every Korean higher education institution or research organisation operating within that law:

  • Articles 307/310 (criminal defamation): true facts that damage reputation are criminal even if true; exemption requires proving the statement is true and solely for the public interest, a defense determined by Korean courts
  • Article 297 (rape): requires "violence or intimidation" rendering the victim unable to resist; nonconsensual sex without physical violence is not legally cognizable as rape in Korean courts

The EU GEP framework's operational requirements — qualitative data collection, evidence-based training, functioning GBV reporting — each embed premises about the national legal environment that Korean law does not satisfy. The result: every institution in the affected jurisdiction faces the same compliance ceiling, regardless of institutional intent.

Three Fault Lines

Fault Line A — Data Collection (Mandatory Requirement #3)

The EC guidance specifies that GEP data collection must include qualitative methods: climate surveys, focus groups, interviews, and — explicitly — sexual harassment data as the paradigmatic gap that GEPs must fill.3 A focus group participant who identifies a named colleague's conduct is simultaneously providing the evidence the EU framework requires and potentially committing a criminal act under Article 307, whose public interest exemption is determined by Korean courts retroactively. EU confidentiality guarantees do not override Korean criminal jurisdiction.

The result is structural sanitization: participants describe experiences in general terms, confirm that problems exist without identifying who is responsible, speak about patterns without naming instances. The data is formally qualitative. It is substantively empty. The GEP built on it addresses a legally permissible description of the institutional environment — not the environment itself.

Sexual harassment data — the specific category the EC guidance identifies as most likely to be absent from administrative records — is the category where Article 307's effect is most direct. Dongguk University's GEP measures its GBV response with two indicators: awareness campaign counts and training completion rates. Neither requires collecting a single item of incident data.2

Fault Line B — Training (Mandatory Requirement #4)

The EC guidance requires that unconscious bias training "identify how institutional processes may cause disadvantage to women" and be grounded in "an evidence-based assessment of the needs of an organisation."3 Training cannot identify institutional processes without a factual base in the institution's own qualitative data. If that data has been structurally sanitized at the collection stage by Article 307, training cannot be evidence-based in the sense the framework requires — because the evidence that would make it so has been legally suppressed before it reaches the training designers.

Fault Lines A and B are mechanically linked. Article 307 prevents specific, attributable testimony → the absence of that testimony prevents evidence-based institutional assessment → the absence of assessment prevents training that identifies specific institutional processes → the training identifies nothing → completion rates go up → the requirement is formally satisfied. Nothing changes.

Dongguk's training oversight is assigned to the ESG Committee chaired by the University President — the same institutional hierarchy that controls faculty contracts, research funding allocation, and appointment decisions.2 Effective unconscious bias training would need to identify that hierarchy as causally implicated in the 18.1% women in leadership and 14.4% women in research funding that the GEP's own diagnostic data discloses. The committee evaluating whether that training meets its targets is the subject of the investigation the training is supposed to conduct.

Fault Line C — GBV Measures (Recommended but Not Optional in Practice)

The EC guidance requires "clear, visible, and robust reporting channels" that ensure potential victims "are not deterred from reporting instances," and "guidance and support for reporting to the police, and legal proceedings against those suspected abusers or harassers, including court cases."3

Korean law forecloses this requirement at both levels simultaneously.

At the legal proceedings level: Article 297 means that the conduct constituting the majority of campus sexual violence — documented in the National Sexual Violence Relief Centre's 2019 survey as present in more than two-thirds of cases processed by counseling centers — is not legally prosecutable as rape. An institution directing survivors toward police reporting without disclosing this is not offering a genuine pathway to justice; it is directing them toward a system that will not recognize what happened to them as a crime.4

At the reporting level: Article 307 means that a survivor who files a formal complaint naming a perpetrator — and whose complaint is later attributed to her — may face criminal prosecution for making a true statement that damages the perpetrator's reputation. Korean MeToo accusers have been prosecuted under this statute since 2018. Human Rights Watch documented in 2025 that the Yoon government used criminal defamation at higher rates than its predecessors.5 A reporting channel that creates criminal counter-prosecution exposure for its users is not non-deterrent. It is structurally deterrent by operation of law.

An institution cannot offer both a genuinely non-deterrent GBV reporting mechanism and accurate legal information about the risks of using it. The EU framework's requirement assumes a national legal environment where reporting is protected. Korean law inverts that protection at both ends of the reporting pathway.

The Case Study: Dongguk University

Dongguk's GEP is the predicted output of the three fault lines — produced by a specific institutional crisis, exhibiting exactly the shape the structural analysis predicts.

Six acts. Ten months. Zero vertical accountability.

Act Date What happened
Legal threat May 2025 Sidus FNH (co-located entertainment firm, run by bid-rigging convict installed as Dean in 2023) issued retraction demand within 24 hours of GW's sexual violence documentation
Tokenism July 2025 Female faculty profile silently edited; caught in real time by Visual Ping automated monitoring
Domestic containment By Sept 2025 Dean Tcha Seung-jai removed from Korean-language faculty page without announcement
Panic scrub Jan 2026 UBC deleted from Global Partners registry; Toronto Metropolitan University reverted to dead name "Ryerson" — caught mid-execution; exposed underlying fraud: list drawn from expired databases
Presidential GEP March 2026 GEP signed by President, task force formed December 2025 — reactive timing anchored to GW's 34-partner fraud exposure going international
Faculty purge April 1, 2026 Two female research visiting professors removed; BK21 government credential eliminated from roster; senior film professor stripped of all institutional contacts — eight days after EU-Korea Research Summit

Every act targets the institutional image. Not one act engages the institutional problem.

What the GEP's own data proves: women hold 18.1% of leadership positions and 14.4% of researchers in internal funding programs.2 The Korean Women's Development Institute's 2020 national study — publicly available five years before the GEP was written — documents 61.5% sexual violence prevalence among female students in Korean arts and culture university programs, with film and media departments carrying an 81/100 structural risk score based on Gender Watchdog's analysis of the KWDI data.6 The GEP's diagnostic section cites the institutional averages. It does not cite the KWDI study. It does not name the Graduate School of Digital Image and Contents once.

The EU Delegation's acknowledgment: On March 19, 2026, Counsellor for Digital and Research Rainer Wessely confirmed in writing that Gender Watchdog's GEP compliance briefing had been shared with RTD units dealing with international cooperation and gender equality, and that the Delegation would "verify whether any of the proposals under Horizon Europe calls are not in conformity with the GEP requirements." The faculty roster cleanup ran eight days later.7

The Jurisdiction-Wide Dimension

Dongguk is not an anomaly. It is a data point.

The 11/12 finding spans institutions with radically different profiles: SNU (Korea's most prestigious research university), KAIST (global top-50 STEM institution), Ewha Womans University (Korea's most prominent women's university, home to a Graduate School of Gender Studies), IBS, SKKU, and six others. These institutions do not share a common administrative culture, a common funding profile, or a common history of institutional crisis. They share a legal environment. The common variable that explains the common output is Articles 297, 307, and 310 of the Korean Criminal Act.

The most structurally significant absence: Chung-Ang University's professor holds the National Contact Point role for Horizon Europe Clusters 1 and 6 — the official Korean government-designated liaison for Horizon compliance guidance. His home institution has no GEP.1 The compliance chain begins at an institution with a compliance gap.

Six Recommendations

1. EC DG RTD — Commission a formal legal compatibility assessment of Korea's Criminal Act (Articles 297, 307, 310) against the operational premises of the four mandatory GEP requirements, before treating Korean GEP filings as presumptively valid compliance evidence.

2. EC DG RTD / REA — Develop a legal environment checklist for Associated Country GEP compliance. Where a country's national law does not satisfy the checklist conditions, introduce a conditional compliance status that flags the structural limitation to proposal evaluators.

3. EU Delegation Seoul — Brief GEP verification staff that Article 307 criminal exposure persists regardless of interview location, that absence of specific testimony is the expected output of a legally constrained environment and is not evidence of institutional compliance, and that "no incidents identified" findings from Korean contexts must be annotated with this legal environment note.

4. EIGE — Add a legal environment assessment module to the GEAR toolbox for Associated Country contexts. Document the Korea case — and Dongguk's GEP — as the first formally documented instance of a national legal architecture structurally incompatible with the framework's operational premises.

5. EC DG RTD / NRF / K-ERC — Issue a 90-day GEP publication requirement to Chung-Ang University as a condition of its professor retaining the NCP designation for Clusters 1 and 6. Establish GEP compliance of the appointee's home institution as a minimum eligibility condition for the NCP role going forward.

6. EU Delegation / EEAS — Table the four legal reform conditions (Article 307 truth-defense amendment; Article 310 proportionality replacement; Article 297 consent centering; ILO C.190 ratification) as a formal agenda item in the next EU-Korea Human Rights Dialogue, explicitly linked to Korea's Horizon Europe Associated Country obligations. The bills already exist in the National Assembly; the EU has standing to request expedited passage.

Gender Watchdog is supported by EROC (End Rape On Campus).

Key Sources

  1. Gender Watchdog, "GEP Theatre and the Unguarded Gate" (March 18, 2026) — 11/12 finding, CAU NCP, K-ERC forum dataset, EURAXESS workshops. https://blog.genderwatchdog.org/gep-theatre-dongguk-chung-ang-horizon-europe-unguarded-gate/ ↩ ↩2

  2. Dongguk University, Gender Equality Plan (archived March 19, 2026). https://web.archive.org/web/20260319034450/https://www.dongguk.edu/eng/page/1173 ↩ ↩2 ↩3 ↩4

  3. European Commission, DG RTD, "Horizon Europe Guidance on Gender Equality Plans" (September 2021). ISBN 978-92-76-39184-5. https://doi.org/10.2777/876509 ↩ ↩2 ↩3

  4. Human Rights Watch, "South Korea Cancels Plans to Update Definition of Rape" (February 1, 2023). https://www.hrw.org/news/2023/02/01/south-korea-cancels-plans-update-definition-rape ↩

  5. Human Rights Watch, "World Report 2025: South Korea." https://www.hrw.org/world-report/2025/country-chapters/south-korea — Korea Economic Institute of America, Sang Hyun Back, "Problems with Korea's Defamation Law" (January 18, 2019). https://keia.org/the-peninsula/problems-with-koreas-defamation-law/ ↩

  6. Korean Women's Development Institute (KWDI), "Current Status of Sexual Violence Against University Students in the Culture and Arts after the Me Too Movement and Policy Issues" (2020). https://eng.kwdi.re.kr/inc/download.do?ut=A&upIdx=102748&no=1 ↩

  7. Gender Watchdog, "Two Profiles, Two Cleanup Tracks: Dongguk Scrubs Its Film Faculty Eight Days After the EU-Korea Research Summit" (April 1, 2026). Wessely correspondence on file. https://blog.genderwatchdog.org/dongguk-faculty-purge-paper-faculty-eu-cleanup-april-2026/ ↩