Does Google’s Veo 3 Do Enough to Distinguish AI and Reality?
Google has introduced visible watermarks to videos generated by its latest AI model, Veo 3, following its release in May.
The tech giant announced the change through Josh Woodward, Vice President of Google Labs and Google Gemini, who confirmed that all Veo 3 videos will now carry a small ‘Veo’ watermark in the bottom right-hand corner.
The watermark appears as pale white text that can be challenging to spot during casual scrolling through social media feeds, but is visible upon close inspection.
The watermark is set to apply to all Veo 3 videos, except those generated using Google's Flow tool by users with a Google AI Ultra plan.
This marker comes in addition to Google's existing SynthID watermark, which is invisibly embedded in all of the company's AI-generated content, helping to unequivocally identify Veo’s handiwork.
Google has also developed a SynthID detector, though this tool is currently only available to early testers and has not yet been released to the general public.
Is Veo 3’s realism a cause for concern?
Since its introduction at Google I/O 2025, Veo 3 has attracted significant attention for producing exceptionally realistic videos complete with audio and dialogue.
The AI model has been used to create everything from fantastical scenarios — such as the Loch Ness Monster playing the bagpipes — to more mundane content including mock interviews, influencer advertisements, fake news segments and product unboxing videos.
While trained observers can identify telltale signs of AI generation (such as overly smooth skin textures or warping in the images), people are concerned that casual viewers scrolling through social media may not notice these subtle indicators, especially those who are not digital natives.
Why experts are concerned about the effectiveness of the watermark
Digital Forensics Expert Hany Farid has expressed scepticism about the watermark's visibility, noting that "this small watermark is unlikely to be apparent to most consumers who are moving through their social media feed at a break-neck clip".
Elsewhere, Negar Kamali, Researcher at the Kellogg School of Management studying AI content detection, suggests that the watermark must be more prominent.
"Either the watermark needs to be more noticeable, or platforms that host images could include a note beside the image — something like 'Check for a watermark to verify whether the image is AI-generated,'" she says.
Both experts acknowledge that visible watermarks can be easily removed through cropping or editing, though Google's invisible SynthID watermark is considered more resilient to tampering.
The challenge of detection
The introduction of watermarks represents what Google describes as "a first step" whilst the company works to make its SynthID detector more widely available.
However, the solution highlights the ongoing challenge of helping consumers identify AI-generated content without requiring specialised detection tools.
As Hany says, whilst the invisible SynthID watermark is difficult to remove, "the average user can't see this watermark without a watermark reader".
When AI video becomes indistinguishable from reality
As AI video generation technology rapidly advances, we're approaching a threshold where synthetic footage will become completely indistinguishable from authentic recordings.
Current models like Veo 3 already produce remarkably convincing content, with remaining telltale signs diminishing with each iteration.
The arrival of truly indistinguishable AI video could fundamentally alter how society processes visual information and establishes truth.
“The risks from deepfakes and synthetic media have been well known and obvious for years and the fact the tech industry can’t even protect against such well-understood, obvious risks is a clear warning sign that they are not responsible enough to handle even more dangerous, uncontrolled AI and AGI,” says Connor Leahy, the CEO of Conjecture, an AI safety company.
“Companies need to be creating mechanisms to distinguish between authentic and synthetic imagery right now,” Margaret Mitchell, Chief AI Ethics Scientist at Hugging Face, told TIME Magazine.
“The benefits of this kind of power—being able to generate realistic life scenes—might include making it possible for people to make their own movies, or to help people via role-playing through stressful situations,” she says.
Google has stated that it plans to focus efforts on protection.
In a statement delivered to Mashable, a Google spokesperson said: “Veo 3 has proved hugely popular since its launch. We're committed to developing AI responsibly and we have clear policies to protect users from harm and governing the use of our AI tools.”
Explore the latest edition of Technology Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
Technology Magazine is a BizClik brand
