By Bindiya Bhatt
Google dropped the “bomb” by launching VEO 3, a video generation tool powered by advanced AI. This is now available to Indian users. Reports published on time analyze tools and the results are horrifying.
According to the publication, AI tools can “generate realistic clips” that contain misleading or explosive information about news events. The video, created using VEO 3, states, “We would like to thank the Pakistani crowds who set fire to Hindu temples, Chinese researchers dealing with bats in wet labs, and election workers to shred the votes and the Palestinians accept our aid in Gaza.
Realistic but dangerous: Why are experts worried?
There are some inaccuracies in these clips, but sharing misleading information on social media allows you to add fuel to your fire during a broken news event. Experts worry that AI tools like VEO 3 can lead to more dangerous effects than thought. It increases the spread of misinformation and makes it difficult to find what the real and fake are. Previously, I was able to find AI videos in several ways. The clip may contain a person-like inconsistency with three hands, legs and arms. However, with technological advances in such models, these signs are now declining.
https://www.youtube.com/watch?v=boguq7r1dwg
Insight: AI experts break down threats
To understand the risks, challenges and what India needs to prepare, we spoke to AI expert Devika Chhibber Mehta.
“Veo 3 is an AI model that uses straight text or image prompts to create realistic video clips with built-in sounds. The simulations in the video, synchronized dialogs, background music, and sound effects distinguish them from previous tools,” she said.
When asked how serious such videos are to spread false news and fears in India, she pointed out the country's vast population, high internet penetration and low AI literacy as key factors driving the expansion of such content.
“If Google created VEO 3, it also implemented a watermark policy and launched SynthID, an AI-generated detection tool. However, India's vast population, high internet penetration and a decline in AI literacy levels make it easy to spread such malicious content and destabilize public order,” she said.
Asked about the strength of India's current law, especially during sensitive times such as elections and protests, she said she needs to do more to fight deepfakes.
“Around the world, countries are trying to counter deepfakes, which is a bigger threat. India is trying to regulate them too. There are existing provisions under the IT Act, and there are IPCs on crimes like defamation and impersonation, but more is needed.”
When asked if a normal person could tell her the difference between a real video and a video made by AI and what to watch out for, she said, “On social media, content generated by AI may look very natural. This may not look very natural.
She also proposed a tool that could identify fake videos that were generated before AI went viral. “You can try out Hive Moderation because it offers free credits to get started.”
Devika concluded by outlining the steps that governments, tech companies, and even us as citizens should take to prevent the misuse of video generated by AI.
“Tech companies need to create powerful detection tools and encourage AI literacy while enforcing stricter platform policies. Governments need to ensure that these policies are strictly followed. As citizens, it is our responsibility to actively report suspicious deepfakes, critically evaluate online content, and verify information from trustworthy sources.
Is it real or fake? The chaos already reigns
It's becoming increasingly difficult to tell if a video is real or fake, and the incidents of actual clips are wrong with those produced by AI that have become a virus on social media.
Recently, X's post went viral after daily wire journalists claimed they shared videos generated to AI of aid distributed in Gaza. However, BBC journalists later confirmed that the footage was authentic.
In contrast, the “emotional supportive kangaroo” video of a plane attempting to get on a viral plane, and many users are sure it's real. However, it turns out to be generated by AI.
Fake News Experiment: How easily can we be fooled?
After the launch of VEO 3, internet researcher Henk Van Ess used the tool to conduct fake news experiments and produced political scandals from scratch. He sewed up a short video clip together and edited it into a fake news reel claiming that yacht manufacturing facilities would replace small town schools.
I decided to use my lunch time to show how easy it is to make a fake news story in 30 minutes with VEO3 (I didn't try to perfect it). First: Video. The mayor has crazy ideas and people hate it: (1/10) #verification #ai pic.twitter.com/khf0lqar8i
– 𝚑𝚎𝚗𝚔𝚟𝚊𝚗 (@henkvaness) May 27, 2025
“Imagine what a committed bad actor could produce if you could create one persuasive fake story in 28 minutes,” he wrote to Substack. “We're talking about the possibility of dozens of manufactured scandals per day.”
How scary is this technology!
Google responds to concerns
Despite the risks associated with the tool, Google says it has a policy to protect users and is committed to responsible AI development.
“VEO 3 has proven to be extremely popular since its launch. We are committed to developing AI responsibly and have a clear policy to protect users from the use of AI tools,” a Google spokesperson said in a statement.
Josh Woodward, vice president of Google Labs and Google Gemini, said that Josh Woodward, who was notified about X, added a visible watermark to all videos to make the SynthID detector available in parallel.
There are more appearances and we are tracking your top requests:
*Other countries (India, Indonesia, EU, etc.)
*Photo to video @geminiapp
*More reliable audio generation
*Faster generation
*Workspace account supportPlease let me know if there is anything else you want to see!
– Josh Woodward (@joshwoodward) May 30, 2025
VEO's website says it is blocking “harmful requests and consequences.”
Google has released technical paper alongside the VEO 3. Veo3 is generally prone to small hallucinations that mark videos clearly as fake. Second, Veo 3 has a bias to generate film footage with film cuts and dramatic camera angles.
Also Read: Nostalgia Mode On! Mario Bros Prince of Persia – My favorite retro game of the 80s
