Sensation or Viral Misinformation?
In the last week, a disturbing video has gone viral on social media. The video seems to show a terrible incident wherein a woman is violently attacked by a dolphin (or orca) in front of a shocked audience at a dolphin show, who appears to be “Jessica Radcliffe” according to texts in the video.
Moreover, the edited clip was quickly shared across TikTok, Facebook, X (formerly Twitter), and Instagram Reels, going viral within days and racking up millions of views. Accompanying the clip were captions claiming it was “real footage” of a terrible marine park accident.
However, after investigating, PresentPakistan uncovered a whole different truth: the supposed “Jessica Dolphin Accident” is a fabrication. The clip is a non-real video, and the accident didn’t happen; instead, it is an AI-generated hoax that used advanced video synthesis technologies and was shared to provoke audience reaction.
Table of Contents
- The Origins and Spread of the Video
- Day-by-Day Spread Timeline
- Day-by-Day Spread Timeline
- Verifying the Video
- Reverse Image Search
- Audio Forensics
- Credible Media Cross-Check
- Reverse Image Search
- Indicators of AI Generated Video
- How are these types of videos generated with AI?
- Script & Storyboarding
- AI Video Generation
- Voice Cloning
- Post-Production Editing
- Script & Storyboarding
- Reasons Why Viral Misinformation Entices People
- The Threat of Trusting AI Hoaxes
- Identifying AI Generated Video
- Conclusion – Lessons from the Jessica Dolphin Hoax
- Frequently Asked Questions (FAQ)
- People Also Search
The Origins and Spread of the Video
Our timeline reconstruction captures how rapidly misinformation can spread.
Day 1- TikTok Post
The first postable video comes from an anonymous TikTok account (no prior posts with news-type content) captioned “Tragic dolphin show accident: Jessica Radcliffe’s last moments.” Because of its shocking premise, it was reposted quite quickly.
Day 2 – Twitter/X Reposts
Multiple meme and clickbait accounts reposted the video with captions like “this is horrifying” and “RIP Jessica.” Consequently, the hashtag #JessicaRadcliffe began to trend
Day 3 – Facebook and YouTube Shorts
Low-quality reposts were showing up on Facebook pages that mostly share viral clips or conspiracy-style content. At the same time, YouTube Shorts accounts were using it as “reaction” content.
Day 4 – WhatsApp and Telegram
The clip entered a private group chat and messaging stream format, thus preventing any possible contingency of viewers checking it for legitimacy.
By day 5, the search terms “Jessica Dolphin Accident” were appearing as trending Google searches in multiple countries.
Verifying the Video
PresentPakistan undertook a three-part verification method to validate the video:
A. Reverse Image Search
We removed key frames from the video and conducted the search using Google Images and TinEye in Comparison.
– Result: We did not find any matches to legitimate news photography or documentation of the marine parts incident with this query.
B. Audio Forensics
We examined the soundtrack using audio analysis packages (audacity).
– Result: The crowd noise had the typical “looping” defect you’d see in stock audio.
– The trainer’s voice appeared to have unrealistic pitch distortions—a common feature of AI voice synthesis.
C. Credible Media Cross Check
We contacted marine parks in the regions related to the captions in the video, and there were no known incidents of this nature.
NDTV, The Express Tribune and The Economic Times fact-checking teams have similarly verified this clip to be AI-generated.
Hopefully this editing is okay:
References:
- NDTV Fact Check
- The Express Tribune
- The Economic Times
Indicators of AI Generated Video
Our analyses of the video frames and audio-visual analyses identified several indicators that it was synthetic media:
– Lip-Sync Mismatches: trainer’s mouth movement did not match spoken audio
– Background Errors: background audience members had their hands out of alignment and with extra fingers
– Lighting Variations:stage lighting appeared to have changed angles abnormally from frame to frame
– Splash Artifacts: The dolphin water splash had pixel smearing, a typical artifact seen with AI generated motion.
These are not visible to the naked eye but are meaningful; you can slow-play the video or zoom in on the frame.
How are these types of videos generated with AI?
The processes for developing a convincing fake video are described here:
Step 1 – Scripting & Storyboarding
The creator writes an invented scenario (in this case, the marine park incident) and often uses AI text services (ChatGPT or Claude) to create descriptive components.
Step 2 – AI Video Generation
The creator uses video generation tools like RunwayML, Pika Labs or Sora to create the video’s visuals by giving it prompts like “realistic dolphin show with accident in front of crowd.”
Step 3 – Voice Cloning
The creator creates the dialogue by using an AI voice service like ElevenLabs or Resemble AI, typically by using generic voice cloning to imitate “real” eyewitnesses.
Step 4 – Post Production
The advocate gets stock video from a stock library service like Getty or Pond5, as well as editing using Premiere or Davinci Resolve, adding sound effects and ambiance (like music and crowd noise) to cover up any imperfections with it.
Reasons Why Viral Misinformation Entices People
There is psychology involved:
- Shock Factor: Extreme or horrible content can have an emotional impact.
- Curiosity: They repost it to ask, “Is this real?” without actually finding out.
- Social Proof: They want to share because it was posted by a lot of people, so that in itself is enough for them.
- Platform Algorithms: Content with high engagement will be pushed out regardless of authenticity.
The Threat of Trusting AI Hoaxes
Trusting and sharing such false videos may result in:
- Public Panic: People may unnecessarily avoid public attractions.
- Public Relations Damage: People and organizations mentioned in the hoax may be damaged.
- Real Issues are Dismissed: Actual events will become diminished.
- Fraud Boards get leverage: Scammers can leverage the clip to ask for fake donations.
Identifying AI Generated Videos
These are a few straightforward ideas that can help you identify fakes before you decide to post:
- Stop and Zoom in on Things or the AI Video: Look for distorted hands, faces or backgrounds.
- Observe Shadows & Lighting Patterns: AI videos typically demonstrate inconsistencies in shadows and lighting.
- Listen for Audible Cues: Robotic-sounding speech or awkward pauses can highlight synthetically created speech.
- Check the Original Source: Research the incident on legitimate news sites.
- Reverse Search the Frame: Search for images in a different context.
Conclusion: Learnings from the Jessica Dolphin Hoax
The Jessica Dolphin video is the most credible example of how believable AI images were still enough to fool millions of people in a matter of days. Although the math for a video like this is astonishing, the application of this technology being weaponized to make a genuine threat to veracity and public trust is frightening.
At PresentPakistan, we take all measures towards battling misinformation through critical thinking and verified sources and awareness of the public. The next time you see a shocking clip, have the attitude that “going viral” does not always equal the greatest truth.
Frequently Asked Questions (FAQ)
Q1: Who is Jessica Radcliffe?
Jessica Radcliffe is not a verified public figure associated with any real dolphin or orca attack. In this instance of the viral video, it appears to be a pseudonym created by the poster to create a more intimate and believable event.
Q2: Did a dolphin or orca attack a trainer recently?
No verified news source or marine park has reported a recent incident like this. The video is an AI-generated video and does not represent an actual event.
Q3: How do we know the video is fake?
There are many indications the video is AI generated, such as the disfiguration of background elements, lip-sync mismatching, looping of additional sound from the crowd, and lack of any real news coverage of the incident.
Q4: What is the process of making AI generated video?
Creators have the ability to offer advanced tools to create visuals from text prompts (e.g., RunwayML, Pika Labs, Sora) and then edit visuals with an AI generated voiceover as well as stock audio/video material.
Q5: Why do people share fake videos?
Emotional shock, curiousity and viral behaviors from engagement algorithms prompt behavior to share content without verification.
Q6: What actions should I take if I see a suspicious video?
Before sharing, take some time to pause, verify the source, run a reverse image search, and check credible news sources to confirm the veracity of the claim.
Q7: What are the dangers of believing viral fake videos?
Spreading fake videos can contribute to public panic, be damaging to reputations, and divert people’s attention away from real problems that demand attention.
People Also Search:
Q1: Did the dolphin attack on Jessica Radcliffe really happen?
No. The video is suspected of being created with AI tools, and nothing indicates that this actually took place in reality.
Q2: Is there a video of a dolphin actually killing a trainer?
No, there is not a record of any verified incident in recent history. Marine parks have a history of incidents; however, none have the same context or background as the clip that is going viral.
Q3: How can I tell if a viral video is AI generated?
There are a number of indicators to look for: do the faces move jerkily or outside of the norm of typical facial flexibility? Did you see any scenes with backgrounds that flipped strangely? Did the lighting make sense and match the audio, and was there audio looping that was overly obvious? And as always, check credible news sources yourself.
Q4: Did this dolphin attack occur within the United States?
We cannot see a place in the video that matches any U.S. marine park, and none of the marine parks have reported such an accident.
Q5: Who created the Jessica Radcliffe viral video?
The original creator is currently unknown. The video appeared first on TikTok as an anonymous user
Q6: What is RunwayML,, and how does it relate to the video?
RunwayML is an AI video generation tool that generates photorealistic scenes from text prompts. It is one of the tools that is regularly used to fabricate this type of hoax.
Q7: Why do people make fake accident videos?
Content creators often want views, ad revenue, or followers on social media. Other creators spread misinformation through these works.
