AI & Digital

EU Deepfake Regulation

CSuggestive similarity

Evidence tier

Echo patterns are similarity-based hypotheses, not proof of causation. Open the sources and evaluate the linkage yourself.

Evidence Scorecard

Three lenses — wording, timing, documented link. Not a single proof score.

Suggestive similarity

Semantic

Moderate keyword/concept overlap

Temporal

Same time period / concurrent

Linkage

Explicit citation in Korean source

Multilingual semantic similarity

cosine similarity (method detail)0.827

Computed with intfloat/multilingual-e5-small multilingual embeddings. Display score is a calibrated mapping of the raw cross-lingual cosine (0.827); higher means closer meaning. Reflects semantic similarity, not proof of causation.

Aligned sentences

EU AI Act Article 50 requires providers of systems generating synthetic audio, image, video or text to mark outputs as artificially generated; the DSA adds platform duties on manipulated media.

aligned

EU Article 50 표시 의무와 국내 라벨링·형사 이중 트랙으로 합성 미디어를 규율한다.

Could also be…show

•Parallel domestic policy development is possible.
•Industry or market pressure may have driven similar language.
•A common external shock (e.g., financial crisis, pandemic) could explain parallel adoption.

Timeline of the echo

Same year

First Korean mention2024 deepfake criminal and labeling track

From date field: 2024 criminalization → 2026 AI Act enforcement · 2024 criminalization → 2026 AI Act enforcement

Echo Evidence Brief

Plain-language explanation for first-time readers — side-by-side echo comparison, verbatim primary-source excerpts, and X-discourse discovery context (not tier evidence).

EU published the global instrument first (Same time period / concurrent; 2024 criminalization → 2026 AI Act enforcement). Korean agency then adopted Korean text using the same policy vocabulary. Matched terms: "deep fake" ↔ "인공지능기본법"; "disclose artificially generated" ↔ "라벨링 의무". Read the side-by-side comparison and verbatim primary-source excerpts below.

Echo pathway

01
Global instrument
EU
02
Temporal direction
Same time period / concurrent · 2024 criminalization → 2026 AI Act enforcement
03
Korean adoption
Korean agency

Global agenda precedes Korean statute or agency program — same vocabulary, traceable sources.

Vocabulary bridge (matched concepts)

These term pairs are the visible echo signal — the same policy idea under different languages.

deep fake인공지능기본법disclose artificially generated라벨링 의무

Side-by-side echo comparison

Global (EU)

EU AI Act Article 50 requires providers of systems generating synthetic audio, image, video or text to mark outputs as artificially^② generated; the DSA adds platform duties on manipulated media. Global enforcement discourse (X threads on Brazil–Korea fake-news parallels) flagged this lane before primary-law verification.

Korean echo (Korean agency)

「인공지능기본법^①」: 생성형 AI 산출물 라벨링^②·허위조작정보(딥페이크) 조항 + 과태료. 형법 제14조의2(2024 개정): 딥페이크 성적 영상물 제작·유포·소지·시청 형사처벌. EU Article 50 표시 의무와 국내 라벨링·형사 이중 트랙으로 합성 미디어를 규율한다.

Global source excerpt

EU AI Act (Regulation 2024/1689) · Article 50(2)

Article 50(2): Providers of AI systems, including general-purpose AI models, generating synthetic audio, image, video or text content, shall ensure that the outputs of the AI system are marked in a machine-readable format and detectable as artificially generated or manipulated. Providers shall ensure that technical solutions are effective, interoperable, robust and reliable as far as this is technically feasible.

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R1689

Korean source excerpt

National Law Information Center · Act on Special Cases concerning Punishment of Sexual Violence Crimes

「인공지능기본법」 딥페이크 라벨링 의무 + 형법 개정(2024): 악의적 딥페이크 제작·유포 처벌(법제처·경찰청 지침 확인). 제14조의2 2024 개정은 딥페이크 성적 영상물의 제작·유포·소지·시청을 형사처벌 대상으로 삼아 EU AI법 Article 50 합성 미디어 표시 의무와 국내 형사·라벨링 이중 트랙을 형성합니다.

https://www.law.go.kr/lsInfoP.do?lsId=011187

X / discourse discovery leadsnot tier evidence

[E1, E4]Discourse on deepfake criminalization and DSA manipulated-media duties surfaced Korea’s parallel Article 14-2 + AI Act labeling path.

Interrogate the Evidence

Answers are rendered directly from this record’s own curated fields — nothing is generated.

Csuggestive similarity · 72%

SemanticModerate keyword/concept overlap
TemporalSame time period / concurrent
LinkageExplicit citation in Korean source

Matched concepts: deep fake ↔ 인공지능기본법; disclose artificially generated ↔ 라벨링 의무

Cross-lingual similarity 60/100 (multilingual e5 embedding) — a computed signal, not proof.

Similarity type: platform-content regulation — EU-KR parallel

Caveat: semantic similarity only — this tier records a hypothesis, not proof of causal influence.

Quote bank

Short verbatim anchors from the global instrument and Korean law/policy — not a full stenogram.

Global

EU AI Act Article 50 requires machine-readable marking of synthetic media; the DSA adds platform duties for systemic risks from illegal and manipulated content.
EU AI Act Art. 50 / DSAsource

Korean

생성·합성 미디어 라벨링 의무와 딥페이크 성적 영상 형사처벌 트랙이 병행 — EU 투명성 논의와 국내 형사·표시 이중 규율.
AI 관련 법령 + 성폭력처벌 특례 체계source

Matched Concepts

2 pairs

GlobalKorean Echo

deep fake

↓→

인공지능기본법

disclose artificially generated

↓→

라벨링 의무

Suggested Concepts

1 suggested

Auto-aligned concept pairs surfaced by our Concept Suggester engine. Shown as supplementary leads — the curated matches above are the authoritative set.

GlobalSuggested Korean

EU AI Act Article 50 requires providers of systems generating synthetic audio, image, video or text to mark outputs as artificially generated; the DSA adds platform duties on manipulated media.

Keyword↓

→

「인공지능기본법」: 생성형 AI 산출물 라벨링·허위조작정보(딥페이크) 조항 + 과태료.

Agenda Arc

Curated

Global purpose, Korean purpose, and precursor stages. Reasoning echo is discourse observation — not proof of coordination.

Global purpose

EU AI Act Article 50 requires providers of systems generating synthetic audio, image, video or text to mark outputs as artificially generated; the DSA adds platform duties on manipulated media.

Korean purpose

「인공지능기본법」: 생성형 AI 산출물 라벨링·허위조작정보(딥페이크) 조항 + 과태료. 형법 제14조의2(2024 개정): 딥페이크 성적 영상물 제작·유포·소지·시청 형사처벌.

국가법령정보센터 · 인공지능기본법(딥페이크) — 「인공지능기본법」: 생성형 AI 산출물 라벨링·허위조작정보(딥페이크) 조항 + 과태료. 형법 제14조의2(2024 개정): 딥페이크 성적 영상물 제작·유포·소지·시청 형사처벌.

Reasoning echo

Observed framing link: platform-content regulation — EU-KR parallel. Not a claim of coordination.

Started where

EU · UN

Started how

2024 criminalization → 2026 AI Act enforcement

CapabilityGlobalEU
AI / generative media capability
Pattern text references AI or generative-media capability that later regulation addresses.
Primary source
Soft lawGlobal2024EU
criminalization
2024 criminalization
Primary source
EnforcementKorea2026EU
AI Act enforcement
2026 AI Act enforcement
Primary source

Original (Global)

Echo (Korea)

Echo Strength

sourceVerified

hasKoreanSource

hasGlobalSource

hasDate

keywordDiversity

agendaTags

Origin

EUUN

Timeline

2024 criminalization → 2026 AI Act enforcement

platform-content regulation — EU-KR parallel

Sources

🌍eur-lex.europa.eu archived

🇰🇷국가법령정보센터 · 인공지능기본법(딥페이크)

Linkage Evidence

Concrete actors, citations, and consultations that connect the global origin to the Korean echo.

Citations

EU AI Act · Article 50 transparency (Regulation 2024/1689)

Article 50(2) EU AI Act: Providers generating synthetic audio, image, video or text content shall ensure outputs are marked in a machine-readable format and detectable as artificially generated or manipulated — the same transparency frame Korea mirrors in AI Act labeling plus criminal deepfake provisions.

National Law Information Center · Act on Special Cases concerning Punishment of Sexual Violence Crimes

「인공지능기본법」 딥페이크 라벨링 의무 + 형법 개정(2024): 악의적 딥페이크 제작·유포 처벌. 제14조의2 2024 개정은 딥페이크 성적 영상물의 제작·유포·소지·시청을 형사처벌 대상으로 삼아 EU DSA·AI법의 합성 미디어 투명성·집행 논의와 평행하는 국내 조문 체계를 완성합니다.

Related Patterns

Korea AI Framework Act (2026)

EUUN

P(B|A) = 75%temporal

40%

EU CBAM & Korea Response (2026)

EUUN

P(B|A) = 50%agenda

40%

Platform Content Moderation (Network Act)

EUUN

P(B|A) = 75%temporal

40%

Supply Chain Due Diligence (CSDDD)

EUUN

P(B|A) = 50%agenda

40%

EU–Korea Framework Agreement → Political / Global Governance Cooperation

EUUN

P(B|A) = 35%agenda

40%

EU ETS → Korea Emissions Trading Scheme (K-ETS)

EUUN

P(B|A) = 35%agenda

40%

AI Misinformation Laws

EUUN

P(B|A) = 68%temporal

33%

AI Transparency & Generative AI Labeling

EUUN

P(B|A) = 67%temporal

30%

Semantically Related

Closest patterns by multilingual embedding (e5) — conceptual similarity beyond shared keywords. A computed signal, not a sourced claim.

AI Misinformation Laws

AI & Digital

cos 0.97

AI Transparency & Generative AI Labeling

AI & Digital

cos 0.95

Brazil PL 2630 / Fake News Bill ↔ Korea AI Act Information Offences (Comparative)

AI & Digital

cos 0.91

Korea AI Framework Act (2026)

AI & Digital

cos 0.91

Platform Content Moderation (Network Act)

AI & Digital

cos 0.90

EU AI Act → Korea AI Framework Act

AI & Digital

cos 0.89

Keywords

deepfakesynthetic mediaAI-generated contentlabeling obligationdeepfake detectionmalicious deepfakecriminalization딥페이크합성 영상AI 생성 콘텐츠라벨링 의무형법 개정디지털 성범죄platform liabilityjudicial self-investigationinformation offence

Score Guide

Confidence — Editorial A–D tier from semantic, temporal, and linkage evidence.

Echo Strength — Metadata completeness score. Rewards verified sources, paired links, explicit dates, keyword diversity, and origin tags.

Vocabulary Overlap — Heuristic score measuring shared tokens between global and Korean texts. Reflects loanwords and repeated terms, not meaning across languages.