Skip to content
Submit incident
AII-1001-03881
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

AI-Generated Fake Political Videos Spread During Japanese Election

AI-generated deepfake videos spread on X during Japan's 2026 election, including a video of fictitious elderly women endorsing a politician that reached 1.84 million views. The Japan Fact-Check Center verified the authenticity of disputed videos, while the Grok AI system initially misidentified one video as AI-generated.

28 January 2026 · Chudo Reform Alliance (中道改革連合) · Misinformation and synthetic media
28 January 2026
AII-1001-038812 sources ↓
Quick facts
Category
Misinformation and synthetic media
Country
Japan
Source
朝日新聞デジタル
Date
28 January 2026
Drafted by
Raidu Incident Detection

What happened

Multiple AI-generated deepfake videos spread on X (formerly Twitter) during Japan's 2026 general election. One video posted on January 28 by an account called "Chudo Reform Alliance Supporter [Official]" showed a large crowd at a street speech rally in Fukuoka's Keigo Park and claimed to show real supporters.

Another deepfake video showing fictional elderly women in a street interview saying they supported politician Sanae Takaichi went viral, reaching 1.84 million views and over 20,000 likes as of February 4. The women in the video do not exist and were created with generative AI.

Reported impact

The misinformation claimed real crowds and real supporters, potentially misleading voters about candidate support levels during the election period.

The Grok AI system initially responded to user queries by incorrectly identifying the Chudo Reform Alliance crowd video as AI-generated, contributing to confusion about video authenticity.

Index analysis

The Japan Fact-Check Center found that the large crowd video was not AI-generated, noting that other video and image angles from the same day confirmed that many people actually gathered at the rally. The center stated that Grok can make mistakes in its judgments.

The deepfake video of elderly women supporters was confirmed to be AI-generated by news reporting, though it originated as a YouTube video before spreading on X.

Governance controls

The Japan Fact-Check Center, an NPO that verifies information on the internet, issued fact-check determinations about the disputed videos on February 3.

Classification

Category
Misinformation and synthetic media
Industry
Media, social platforms and marketing
Country
Japan
Harm status
Potential harm
Intent
Unknown
Timing
Post-deployment
Harm types
Public interest
Harmed
General public
Failure modes
Deepfake video, Misidentification
AI systems
Grok

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

朝日新聞デジタル
「早苗ちゃんの演説見た?」拡散した動画、実はAI作成 見破るには:朝日新聞
3 February 2026
nikkansports.com
中道演説会の"大聴衆動画=AI疑惑"拡散に日本ファクトチェックセンターが「判定」発表 - 政治 : 日刊スポーツ
4 February 2026
Aug 2026AI-generated animal images spread online, raising wildlife conservation concernsReportedMisinformation and synthetic mediaFranceinfoSudOuest.frAug 2026Bombay High Court Orders Removal of AI-Generated Deepfake Content Targeting Nitin GadkariConfirmedMisinformation and synthetic mediaBBCPatrika NewsJul 2026Bombay High Court permits Nitin Gadkari to sue over deepfakes linking him to E20 fuelAllegedMisinformation and synthetic mediaThe HinduThe Financial ExpressIndia News, Breaking News, Entertainment News | India.com+5Jul 2026AI deepfakes fueled disinformation during CJP protests in MaharashtraConfirmedMisinformation and synthetic mediaThe Times of IndiaLokmat Timesmid-day+3Jun 2026AI chatbots spread election misinformation and recommend candidates in BrazilReportedMisinformation and synthetic mediaEstadãoDiário da Região