Artificial intelligence is making online scams more convincing, allowing fraudsters to imitate familiar faces, clone voices and create advertisements that appear genuine. A recent warning from the cybercrime
authorities has highlighted how scammers are using manipulated videos of political leaders to promote fake government welfare schemes and trick people into transferring money.
On October 19, the Indian Cyber Crime Coordination Centre (I4C), under the Union Home Ministry, flagged fraudulent advertisements circulating on Facebook and Instagram. These ads feature deepfake videos of Prime Minister Narendra Modi, Union Finance Minister Nirmala Sitharaman and chief ministers, falsely portraying them as endorsing schemes that promise financial assistance to people in need.
The warning highlights a growing challenge for internet users: when a video, voice or message looks and sounds authentic, how can people tell whether the person behind it is genuine? From fake welfare benefits to fraudulent shopping websites and recruitment calls, AI is giving scammers new ways to exploit trust.
How Do Fake Welfare Scams Work?
The fraud often begins with a social media advertisement featuring a familiar political leader announcing a supposed government scheme. Viewers are promised financial assistance, cashback rewards or other benefits, making the offer appear credible because it is associated with a trusted public figure.
Once a person responds, fraudsters may ask for registration charges, processing fees, taxes or other payments before the promised benefit can be released. In reality, the scheme is fake, and the money goes to the scammers.
Some advertisements use scratch cards or spin-and-win contests to offer cashback rewards. Victims may be tricked into approving fraudulent Unified Payments Interface (UPI) collect requests, which can result in money being debited from their accounts instead of credited to them. Authorities have also warned about suspicious calls from international numbers claiming to represent public figures or their offices.
The deception works because it combines the appearance of official endorsement with the promise of financial gain. People who believe they are applying for a legitimate benefit may share personal information or make payments before checking whether the scheme exists.
A familiar face, however, is not proof of authenticity. A video featuring a political leader does not establish that the person endorsed the offer or that the scheme has government approval.
“Fraudulent welfare advertisement messages mirror authorities’ notifications about financial subsidies, cash transfers and services at no cost. Imposters make use of logos that look authentic, non-existing references and narrow deadlines to gain trust. The next step is leading the victims to a fake website or WhatsApp link, where they provide their personal data and banking information in order to register, pay for the service or confirm the transaction,” said Mandar Patil, Executive Vice-President at Cyble — an AI-native cybersecurity company.
Beyond Deepfakes: How Scammers Deceive People
Scammers can use AI-generated images, convincing text, cloned voices and fabricated identities to make fraudulent offers appear legitimate across several online channels.
Fake shopping websites, for instance, may imitate established retailers, use polished product images and advertise unusually attractive discounts. A buyer who pays in advance may receive counterfeit goods, an inferior product or nothing at all. Some websites are designed primarily to collect payment details and personal information.
Jobseekers face a different risk. Fraudsters can pose as recruiters or human resources representatives, using convincing company branding, fabricated job descriptions and professional-sounding calls to gain trust. They may then demand registration fees, interview charges or payments for training and background checks. Legitimate employers generally do not require candidates to pay for a job offer.
Voice cloning creates another route for impersonation. A scammer may imitate the voice of a relative, friend or colleague and claim to need urgent financial assistance. The request may arrive with a fabricated emergency designed to discourage the recipient from checking the story.
Investment scams can use similar techniques, with fake endorsements, fabricated testimonials and promises of unusually high returns. The common thread is not the technology itself, but how it is used to make an offer or identity appear trustworthy.
“Scammers seem to be taking advantage of AI-generated deepfakes and V-shaped voices, and launching phishing campaigns that use credible websites for the purpose of entrapment. Their high level of creativity allows them to imitate employers, financial institutions, retailers, and brands through clever click bait such as special discount offers, job offers, and alerts about urgent issues with one’s account. Most importantly, for businesses, there is a threat because fraudulent activity in imitation of their identity destroys their brand image and erodes public trust in them,” said Hakimuddin Wadlawala, Founder, Aquila I – a AI-native cybersecurity platform.
Why Are AI-Powered Scams Harder To Detect?
Traditional online scams often contain clues such as spelling mistakes, awkward wording, suspicious links or poorly designed websites. GenAI could help criminals produce fluent messages, realistic images and more polished advertisements, reducing some of these obvious warning signs.
Deepfake technology can also manipulate a person’s facial movements, expressions and voice to create the impression that they are speaking on camera. Although some manipulated videos still contain visual or audio inconsistencies, these are not always obvious to an ordinary viewer, particularly on a small phone screen.
The greater danger is the combination of technology and social engineering. Scammers create urgency by claiming an offer is available for a limited time, a payment must be made immediately or a benefit will be lost unless the recipient acts. When a message appears to come from a government leader, a bank, an employer or a family member, people may be less inclined to question it.
AI does not need to create a perfect imitation to succeed. It only needs to make a claim believable enough for someone to click a link, share sensitive information or approve a transaction.
“It is becoming harder to define scams because criminals are able to compose grammatically correct texts, use realistic voices and create videos of excellent quality as well as operate mass-mailing campaigns. It becomes difficult to rely on traditional features of fraud such as spelling mistakes and suspicious profiles as well. Scammers are effectively using the mechanism of trust both in the platforms and in relationships with people, thus making the process of fraud easier,” said Patil.
Why Digital Users Need To Be Alert
India’s expanding digital payment ecosystem has made it easier for consumers to transfer money, pay bills and shop online. However, the convenience of instant transactions also means that a fraudulent payment can happen within seconds.
“Web shoppers are susceptible to buying counterfeits, being victims of payment scams, or having their credentials stolen from fraudulent retail portals. Job hunters might also fall prey to scammers impersonating recruiters and asking them to pay registration fees or submit personal papers. UPI product users might have to deal with counterfeit payment confirmations, collect requests, and social engineering frauds. Other than targeting individuals, these threats can also jeopardise businesses when compromised employee accounts lead to data breaches in business applications, customer data, or payment systems,” said Wadlawala.
The scale of the challenge is reflected in data cited by EY India. Around 68% of cybercrime complaints recorded by the National Crime Records Bureau in 2022 were linked to online financial fraud. EY also noted that the value of money siphoned off in bank frauds nearly tripled in 2024-25, underscoring the growing importance of fraud detection and prevention.
The risks extend across age groups and income levels. Welfare scheme applicants may be attracted by the prospect of financial assistance, shoppers by deep discounts, jobseekers by employment opportunities and investors by promises of quick returns. Digital familiarity alone does not guarantee protection when a scam is designed to exploit trust and urgency.
The response is also evolving. Financial institutions and businesses are increasingly using AI to identify unusual transaction patterns, flag suspicious activity and detect potential fraud in real time. But these tools need reliable data, effective monitoring and human oversight to be useful. Technology can help detect fraud, but it cannot eliminate the risk entirely.
How Can You Protect Yourself From AI Scams?
The most effective defence is to verify the claim independently rather than trust the face, voice or branding used to promote it.
If an advertisement promises government money, check the relevant ministry’s official website or verified government channels. Do not rely on a video circulating on social media as proof that a welfare scheme exists. Be particularly cautious if you are asked to pay a processing fee, registration charge or tax to unlock a promised benefit.
“Make sure to validate any welfare schemes by referring to the respective government official websites before making the application. Never send any cash in advance for promised benefits and jobs without verification. If buying something online, check the URLs and sellers before making the purchase, and refrain from entering any UPI PINs in order to get cash. Verify all urgent requests received from friends, family and employers. Avoid any unknown links, activate transaction alerts, make use of multifactor authentication and report any fraud immediately by calling 1930 helpline in India,” says Patil.
For UPI transactions, read the payment request carefully before approving it. A collect request is not a payment you are receiving; approving it may authorise money to be debited from your account. Never share your UPI PIN, passwords or one-time passwords with anyone claiming to help release funds, said Wadawala.
“For organisations, additional safeguards include enabling multi-factor authentication, training employees to identify AI-driven phishing attempts, strengthening incident reporting mechanisms and tracking impersonation cases. Organisations should also continuously monitor fake websites, lookalike domains and fraudulent social media profiles, supported by clearly defined incident reporting and response procedures. Combining regular employee awareness, robust technical safeguards and proactive threat monitoring can help reduce the risk of financial fraud, credential theft and reputational damage,” he added.
If a friend, relative or colleague contacts you with an urgent request for money, call them back on a number you already know. For suspicious job offers, contact the employer through its official website or recruitment channel. When shopping online, check the seller’s credentials, return policy and payment options before placing an order.
If you suspect financial cyber fraud, report it immediately by calling India’s national cybercrime helpline, 1930, and file a complaint through the official National Cyber Crime Reporting Portal at cybercrime.gov.in. Quick reporting can help authorities and banks take steps to address fraudulent transactions.
Ultimately, AI is changing how scams are created, but the underlying tactic remains familiar: exploiting trust to obtain money or personal information. As fabricated videos and voices become more convincing, independently verifying a claim before acting on it is becoming an essential part of staying safe online.
