Deepfake Security Startup Neural Defend Wins Top Honours at a16z New York Tech Week; Judges from OpenAI, Amazon
New Delhi [India], July 16: At a time when AI-generated deepfakes are challenging digital trust across financial services, governments, and online platforms, an Indian startup is making its mark on the global stage. Neural Defend, which builds real-time deepfake detection technology, has won the Everything AI Startup Pitch Competition at New York Tech Week, one of the largest tech gatherings in the United States, presented by Andreessen Horowitz (a16z).
Selected as the only Indian startup among ten global finalists, Neural Defend secured first place and approximately US$50,000 in equity-free prize funding. The company’s proprietary deepfake detection technology was recognised by an eminent judging panel comprising Derek Canton, Startup Director at OpenAI; Amrita Sankar, North America Head of GenAI Startup GTM at AWS; alongside investors and technology leaders from FirstMark, Amazon’s Alexa Fund, IBM, Geek Ventures, and the Techstars ecosystem.
The recognition comes at a pivotal moment for the AI industry. As generative AI becomes increasingly accessible, deepfakes are evolving beyond misinformation into sophisticated tools for financial fraud, identity theft, and cybercrime. This growing threat is prompting governments, regulators, and enterprises worldwide to strengthen their defences, with AI-powered authenticity verification emerging as a critical safeguard for digital identity, financial systems, and online transactions.
The development is particularly relevant for India, one of the world’s largest digital economies, where secure remote identity verification underpins more than 150 million eKYC verifications annually and billions of monthly UPI transactions. Recognising the growing threat of AI-generated fraud, the Ministry of Home Affairs, through the National Cybercrime Threat Analytics Unit under the Indian Cyber Crime Coordination Centre (I4C), issued an advisory in June 2026, urging banks and fintech companies to adopt advanced deepfake detection technologies after warning that fraudsters were using synthetic media to bypass facial authentication, liveness checks, Video-KYC, and account recovery systems.
The advisory followed the February 2026 amendments to the Information Technology Rules, which introduced stricter obligations around the detection and labelling of AI-generated content while significantly reducing timelines for its removal, reflecting the increasing importance of trusted AI security solutions across the country’s digital ecosystem.
Piyush Verma, Co-founder and CEO of Neural Defend, said, “This recognition at New York Tech Week is a proud moment for us and reinforces our conviction that breakthrough AI technologies can be conceived, built, and scaled from India for the world. It also comes at a defining time, as governments and enterprises are recognising the growing risks posed by AI-generated deepfakes and taking concrete steps to strengthen digital trust. We believe this is just the beginning of a much larger shift, where real-time authenticity verification will become a fundamental requirement across digital ecosystems, and we are committed to building that capability.”
Against this backdrop, Neural Defend is developing technology designed to help enterprises stay ahead of this evolving threat landscape. Its patent-pending detection foundation model is purpose-built to identify AI-generated manipulation across video, images, audio, and live streams in under a second. Unlike many AI startups that build applications on top of existing models, the company is developing its own foundational technology specifically for deepfake detection. Today, its platform is already deployed by leading global financial institutions, including Tier-1 banks, processing millions of detections every month to strengthen customer onboarding, identity verification and fraud prevention.
Katerina Andreeva, AI Strategy Advisor (Ex-IBM) and Competition Judge, said, “Every organisation is beginning to ask the same question: how do we know what we’re seeing is real before we act on it? Neural Defend impressed us because it’s tackling that challenge head-on, with a clear focus on building technology that strengthens trust in AI-driven environments. That level of focus is exactly what sets promising companies apart.”
Alexander Zemlyak – Partner at Geek Ventures, Competition Judge, said, “It was a great event and I really enjoyed it. The quality of the startups was amazing. I’m a fairly frequent judge at different competitions, and this was one of the first times I chose to follow up with most of the startups presenting.”
As artificial intelligence continues to reshape how people communicate, transact, and verify identity, the ability to distinguish authentic interactions from synthetic ones will become increasingly critical. With its proprietary deepfake detection technology and growing enterprise adoption, Neural Defend aims to play a pivotal role in helping organisations navigate the AI era with greater confidence, security, and trust.
About Neural Defend
Built by researchers from MIT and IIT, Neural Defend has developed a proprietary 1-billion-parameter foundation model for detecting AI-generated faces, voices, videos, and documents. Neural Defend was selected into the Bharat AI cohort by NPCI and NVIDIA at Global Fintech Fest Mumbai 2025 alongside Sarvam AI and other top Indian AI startups – where Prime Minister Narendra Modi visited the Neural Defend booth to see the technology firsthand. They also won the Cybersecurity Startup Award at the India AI Impact Summit 2026, and have twice won cybersecurity awards from the Data Security Council of India. It has become the AI detection backbone of India’s digital identity ecosystem, building the authenticity layer for finance: from onboarding and lending to claims, live calls, and devices.
Learn more at neuraldefend.com
If you object to the content of this press release, please notify us at pr.error.rectification@gmail.com. We will respond and rectify the situation within 24 hours.
