Expert shows how to spot a deepfake created with AI
Credibility score: 63/100 — Mostly Credible. Mixed credibility - some claims are solid, others need verification.
BSmeter analyzed "Expert shows how to spot a deepfake created with AI" and rated it 63/100 for credibility (a BS score of 37/100 — mostly credible), on 2026-09-21. Its weakest claim — "Suggesting urgency as a red flag for deepfakes. Good advice, but it's a general scam tactic." — scored 45/100 and was flagged as emotional button. 7 claims were checked against the video transcript. Scores are produced by BSmeter's AI analysis of the transcript, not independent human verification.
Of 7 claims analyzed: 0 scored under 40, 3 between 40 and 69, and 4 at 70 or above.
Claims analyzed
Video opens with a highlight reel preview of the topic: deepfakes and AI. — Just Vibes (50/100)
At 0:05
This is just a quick intro, setting the stage for the 'digital security expert' to talk about AI and deepfakes. Nothing to dissect yet. 😈
Why this score: The segment serves as a brief, attention-grabbing introduction to the video's subject matter—AI and deepfakes—without making any specific claims that require scrutiny. It's pure setup, no substance.
Original quote: “PHASED IN OVER THE NEXT FEW
MONTHS.
>>> ONE OF THE BIGGEST
TECHNOLOGY CONVENTIONS STARTS
IN VEGAS. IT RUNS THROUGH
FRIDAY. YOU COULD BE SURE
ARTIFICIAL INTELLIGENCE WILL BE
A BIG PART OF IT THIS YEAR
>> ONE PERSON FROM THE CHICAGO
AREA WHO IS THERE. DIGITAL
SECURITY EXPERT. BEFORE HE LEFT
FOR VEGAS…”
Defining deepfake as a face swap from phone video. Straightforward. — No Frame (75/100)
At 0:32
Just laying out the basic definition of a deepfake. No tricks here, just the setup. 😈
Why this score: The speaker is simply defining what a deepfake is in a very basic sense, using a common example of a face swap from phone video. There's no hidden agenda or manipulative language; it's a clear, direct explanation to set the stage.
Original quote: “>> AN EXAMPLE OF A DEEP FAKE IS WHEN YOU TAKE VIDEO RECORDED ON A PHONE AND YOU CAN DO A FACE SWAP.”
Claiming the ease and low resources for deepfakes are new. True, but the 'good fake' part is a bit much. — No Frame (75/100)
At 0:40
They're right, the tech is cheaper and faster now. But 'good fake' is subjective, isn't it? 😈
Why this score: The core claim about the increased ease and lower resource requirements for creating deepfakes is accurate. AI tools have indeed made this process more accessible. The 'good fake' part is a slight overstatement, as the video itself goes on to show how to spot them, implying they're not always perfect. Still, the underlying point is valid.
Original quote: “>> Reporter: SCAM ARTISTS HAVE ALWAYS BEEN AROUND BUT WHAT IS DIFFERENT NOW IS HOW EASE TEIS TO MAKE A GOOD FAKE VIDEO OR AUDIO CLIP. >> WHAT HAS CHANGED IS YOU HAD AN ABILITY TO CREATE THE IMAGES BUT THE SPEED AND THE RELATIVE LOW RESOURCES TO DO IT, IT IS DIFFERENT NOW.”
Identifying face/neck completion and face size as deepfake tells. Solid observations. — No Frame (75/100)
At 1:12
Pointing out the obvious tells in the deepfake example. These are real flaws. 🔥
Why this score: The speaker is directly referencing visual cues from the deepfake example shown on screen. The discrepancies in face and neck blending, and the potential for incorrect face sizing, are indeed common artifacts in less sophisticated deepfakes. This is a straightforward observation of evidence.
Original quote: “>> THE COMPLETION ON THE FACE AND THE NECK ARE A BIT DIFFERENT THE SIZE OF THE FACE ON THE FACE SWAP CAN BE OFF A LITTLE BIT AS WELL.”
Highlighting resolution mismatch between face and background as a deepfake sign. Accurate. — No Frame (75/100)
At 1:26
Another clear visual tell. The resolution difference is a dead giveaway. 💀
Why this score: This observation is based on a visible artifact in the deepfake example presented. A mismatch in resolution between the swapped face and the original background is a common indicator of a deepfake, especially when the source image for the face is of higher quality than the target video. This is a direct, verifiable visual cue.
Original quote: “>> IN THIS EXAMPLE YOU SEE ULTIMATELY THE CLARITY OF THE BACKGROUND AND THE FACE ARE VERY DIFFERENT. THAT IS BECAUSE THE IMAGE ON THE FACE IS AT A HIGHER RESOLUTION WHEN IT WAS FACE SWAPPED. IT STICKS OUT.”
Suggesting urgency as a red flag for deepfakes. Good advice, but it's a general scam tactic. — Emotional Button (45/100)
At 2:03
Urgency is a classic scammer move, not just deepfakes. They're hitting that fear button. 🚩
Why this score: While it's true that threat actors often create a sense of urgency, this isn't exclusive to deepfakes. It's a fundamental tactic in almost all forms of social engineering and scams. Framing it as a primary deepfake indicator, rather than a general scam warning, uses an emotional button (fear of being rushed) to connect a broader tactic specifically to deepfakes, potentially overstating its uniqueness in this context.
Original quote: “THERE IS ONE THING THAT SHOULD RAISE A RED FLAG. >> THE THREAT ACTOR IS TRYING TO CREATE A SENSE OF URGENCY. IF YOU FEEL LIKE THE USER IS MAKING YOU FEEL SOMETHING QUICKLY, SLOW DOWN, MAKE THE CHECK”
Technology's wide-ranging impact on kids — a vague, ominous warning. 😈 — Loaded Language (45/100)
At 2:30
He says 'wide-ranging impacts' and 'cost' without naming a single specific impact or cost. That's not information — that's just a vibe. 💀
Why this score: The speaker uses emotionally charged but vague terms like 'wide-ranging impacts' and 'cost' when discussing technology's effect on kids. This creates a sense of alarm without providing any concrete examples, data, or specific problems. It's designed to make you worry without giving you anything to actually worry about. I've seen this trick since before your grandfathers invented 'stranger danger.' 🔥
Original quote: “TECHNOLOGY IT IS HAVING WIDE-RANGING IMPACTS ON A LOT OF PEOPLE AND YOUNGER KIDS AND TEENAGERS, SOME OF THEM IT IS COST”
See the full analysis with timestamps →