Rohan Gupta
Rohan Gupta is the co-founder and CEO of Diopter, which detects deepfake video, cloned voices and AI-driven social engineering on live calls. He previously founded QuillBot and led it through a $200M exit to Course Hero, now Learneo, in 2021. QuillBot today serves more than 35 million monthly users.
From building generative AI to defending against it
Rohan founded QuillBot in 2017 after posting an early version of the tool on Reddit. It became one of the most widely used AI writing platforms and was acquired by Course Hero, now Learneo, in August 2021.
He co-founded Diopter with Jeremy Pippin after seeing how quickly generative AI moved from productivity use cases to impersonation and fraud. Diopter scores live video and voice calls for synthetic media, and for the pattern attackers actually run: authority, urgency, isolation, escalation, and the ask.
Fraud is rarely detectable from a single frame or voice sample. It unfolds across a conversation, and the arc is the unit of analysis.
Track record
AI social-engineering and deepfake detection for live video and voice calls. Detection runs on the endpoint, so only verdicts and scores leave the device.
Built an AI writing platform that today serves more than 35 million monthly users. Raised a $4.25M seed from GSV Ventures and Sierra Ventures, and led the company through a $200M exit to Course Hero, now Learneo, in August 2021.
Recognition and press
Recognised for building QuillBot into a widely used AI writing platform.
On integrating every step of the writing process, and how AI can support language learners.
On the seed round, and on building a decision-making process rather than following best practices.
Articles reviewed
Diopter research checked for technical accuracy before publication.
Caller authentication: How help desks and finance teams check a caller is real
Imagine this situation: a finance executive receives a call from someone claiming to be their regular supplier. The supplier requests an urgent change to their bank account details.…
Are AI detectors accurate? What image, voice, and video detectors get right and wrong
Are AI detectors accurate? What accuracy means, how image, voice, and video detectors perform on real-world content, and why results need other checks.
How to spot fake LinkedIn profiles: check the profile picture for AI signs, verify work history and messages, and use an AI image detector.
In August 2026, McKesson Corporation, a major American healthcare services and pharmaceutical distribution company, disclosed a cybersecurity incident after the ShinyHunters extortion group claimed responsibility for stealing 284…
Six ways to tell if an image is AI-generated: visual checks, fine details, reverse image search, provenance signals, AI image detectors, and context.
The Apollo Global Management data breach in July 2026 shows how social engineering can turn into a cloud security risk. Here is what happened, what was exposed and…
Spear phishing is a targeted form of phishing that uses details about a specific person or organization to make the message seem legitimate. Here's how it works, with…
Vishing turns a phone call into a weapon: scammers use AI voice cloning and caller ID spoofing to impersonate banks, IT staff and executives. Here's how these attacks…
Synthetic identity fraud blends real and fabricated details into an identity that passes traditional checks. How it unfolds, the signals that reveal it, and how to prevent it.
How deepfake attacks run on Zoom, Teams, Meet and Webex, why single-frame checks miss them, and what continuous detection during a live call actually looks at.
How deepfake video detection actually works in 2026: the seven method families, what each one catches, where each one breaks, and how to layer them so a face…
AI social engineering uses AI to make impersonation attacks more convincing and far harder to spot. Here's how the attacks work, why they succeed, and what actually detects…
Help desk scams turn IT support into an entry point, using impersonation to reset credentials and bypass MFA. Here's how the attack runs and how to shut it…
Invoice fraud uses fake invoices, vendor impersonation, and BEC to trigger wrong payments. The main types, a real case, warning signs, and how to prevent it.
Vendor fraud diverts company funds through fake invoices, impersonation, and altered bank details. The main types, two real cases, and the controls that stop it.
Vendor accuracy scores tell you little about production performance. Eleven criteria for evaluating deepfake detection tools, and what to test for each one before you sign.
BEC attacks exploit trust rather than systems. Eight practical strategies to verify requests, harden email authentication, and catch AI-driven voice and video impersonation.
AI agents can plan, decide and act with little human input. That autonomy is also a new attack surface. How agentic AI fraud works, what the first AI-run…
Around 17% of hiring managers have interviewed a candidate using deepfake technology. How AI-generated fake candidates get through hiring, and how to verify who you are really interviewing.
The six detection methods that catch AI social engineering across email, voice, video, and identity, the signals each one reads, and where each one fails.
A forensic breakdown of the 10 deepfake audio detection methods enterprise fraud teams need in 2026: what each one catches, where it breaks, and how to layer them.
AI fraud detection combines machine learning and behavioral analytics to spot suspicious activity in real time, but traditional systems miss scams built on trust. See how AI detects…
Caller ID spoofing lets fraudsters display legitimate bank numbers and pair them with AI-generated voices to sound convincing. Here's how bank impersonation calls work, how to spot the…
Executives are targeted disproportionately by social engineering and deepfake attacks, yet most security stacks were built to defend systems, not people. Here are 8 protection strategies that actually…
Real estate wire fraud targets buyers and sellers during property closings, using impersonation to redirect closing funds to fraudulent accounts. Here's how closing scams work and how to…
Business email compromise (BEC) is one of the most financially damaging cyberattacks organizations face today. Here's how BEC scams work, real-world case studies, and how to protect your…
AI-powered deception has turned social engineering into an assembly line. Here's how enterprise deepfake attacks unfold, why existing security stacks miss them, and what a real prevention program…
Wire fraud is a type of financial fraud that uses electronic transfers to steal money by impersonating a trusted party. Here's how it works and how organizations can…
How a deepfake video call convinced an Arup employee to wire $25 million to fraudsters in 2024, and what the incident reveals about the gaps in traditional verification…
A breakdown of the six major deepfake detection method families, how each one works, where it breaks down under real attack conditions, and what a resilient detection stack…
Social engineering attacks manipulate employees into bypassing security controls entirely. Here's how these attacks work, why they succeed even with strong technical defenses, and how to defend against…
A deepfake is AI-generated synthetic media, audio, video, or images, used to impersonate real people. Here's how the technology works and why it's becoming a serious enterprise security…
Upload a video or audio clip and get an instant AI verdict. No account required. See how Diopter detects synthetic media in seconds.
If you’re dealing with deepfake fraud, synthetic media threats, or need to understand what your organization is actually exposed to, we’d like to hear from you.