As India’s e-commerce boom pulls millions of first-time users online each year, Amazon has issued a pointed warning that generative AI is making cybercriminals faster, more convincing and considerably harder to detect, with the country’s newest digital consumers bearing the brunt of the exposure. The company’s Vice President for Customer and Partner Trust, Rohan Oommen, said India’s fraud landscape looks fundamentally different from mature markets like the United States precisely because its user base is still growing rapidly, and inexperience with digital platforms is proving to be exactly what fraudsters exploit.
A Digital Economy Still Filling Up With First-Timers
Unlike the US, where e-commerce adoption has largely plateaued, Oommen noted that India continues to see a steady influx of first-time digital consumers, many of them from Tier-2 and Tier-3 cities where cyber awareness tends to be lower. That distinction matters enormously for fraud prevention, since scams that a seasoned online shopper would instinctively recognise as suspicious can appear entirely legitimate to someone navigating an app store or a payment gateway for the first time.
This pattern lines up with what India’s own cybercrime data has shown over the past year. The country recorded roughly 25 lakh cyber incidents in 2025, up from 22.68 lakh the year before, with the Ministry of Home Affairs noting that cybercrime has surged by as much as 400 percent since 2021, driven substantially by rural and semi-urban areas emerging as new hotspots for digital fraud. As internet penetration deepens into smaller towns faster than digital literacy programmes can keep pace, the gap between adoption and awareness continues to widen.
Generative AI Cuts Both Ways
Oommen was candid about how generative AI has changed the toolkit available to fraudsters, enabling the rapid creation of fake identities, forged documents and convincingly fabricated reviews that would have taken far more effort to produce even a few years ago. Industry data backs up the scale of this shift, with deepfake files in India projected to grow from roughly 500,000 in 2023 to around 8 million by 2025, a trajectory that reflects how accessible synthetic media generation has become even to low-skill fraud operators using inexpensive, customised AI tools.
Amazon, however, is deploying the same underlying technology defensively. The company is leaning heavily on AI systems capable of continuously monitoring its platform, flagging suspicious activity and running automated risk assessments at scale, moving well beyond document or image verification toward behavioural analytics. Oommen explained that verifying a seller through documents or photos alone is no longer adequate, since forged paperwork can now be generated convincingly within minutes, whereas behavioural patterns built up over months or years of platform activity are considerably harder for a fraudulent actor to fake convincingly.
The Scale of the Cleanup Effort
Amazon’s own enforcement figures give a sense of just how much fraudulent activity large platforms are now filtering out on an ongoing basis. According to the company’s latest Trustworthy Shopping Experience Report, Amazon removed nearly 300 million fake reviews worldwide during 2025, took down approximately 70,000 phishing websites, and acted against roughly 14,000 phone numbers linked to cyber scams. Oommen noted that phone-based fraud remains considerably harder to combat than fraudulent websites, since accurately tracing the individuals behind a phone number involves a different and often slower investigative process than delisting a malicious domain.
He added that Amazon does not wait for regulatory intervention before acting against suspected fraud, describing continuous platform monitoring and immediate response to detected threats as central to maintaining consumer trust, an approach he framed as essential to sustaining confidence in digital commerce as more first-time users come online.
Cybersecurity experts broadly agree that AI-driven fraud will only grow more sophisticated from here, pointing to the rising use of deepfake videos, AI-generated voices, forged documents and automated phishing campaigns as evidence of a threat landscape evolving faster than most consumer awareness efforts. A researcher at Algoritha Security said generative AI has substantially improved fraudsters’ ability to produce convincing fake messages and fabricated identities, making it increasingly difficult for first-time internet users in particular to distinguish legitimate platforms from scams. The researcher recommended that consumers shop only through official websites or apps, avoid clicking unsolicited links, never act on payment requests received through unexpected calls or messages, and enable two-factor authentication wherever available, arguing that no technological safeguard alone can substitute for informed, cautious user behaviour.
