DetectifAI
DetectifAI builds compact, on-device AI models that detect AI-generated and cloned voices in real time during phone calls, without sending audio to the cloud.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
DetectifAI designs small, foundation-model-based deepfake-audio detectors built from the ground up to run directly on a smartphone’s operating system, rather than shrinking a large cloud model to fit on-device. The goal is an instant, on-device verdict on whether a voice on a call, voicemail, or other audio is AI-generated, without the audio ever leaving the phone. The company already processes over 100,000 calls a month for financial institutions in India that use AI voice agents for debt collection and loan follow-ups, pairing deepfake detection with speaker verification on every call.
Founded in 2025 and based in San Francisco and Bengaluru, DetectifAI was started by Tarini Padmanabhuni after a deepfake voice scam targeted her grandfather. The company has raised a small, undisclosed seed amount from angel investors including former TechCrunch editor Josh Constine, and plans to extend its technology to KYC providers, telecoms, and phone manufacturers.
Innovation Matrix Assessment
Founded in 2025, the company already has a live production deployment processing over 100,000 calls a month — unusually fast execution for a pre-seed company.
On-device, low-latency detection is practical for call-center and voice-agent use cases where cloud round-trips are too slow to act on a live call.
Funding is a small, undisclosed pre-seed amount from angel investors; traction is currently concentrated in one sector (Indian financial-services voice agents), so momentum is scored conservatively.
The on-device, purpose-built-small-model architecture is a genuine technical differentiator, but deepfake-audio detection is already an active category with better-funded competitors.
Real, reported production call volume (100,000+/month) is a meaningful operational proof point, though it has not been independently audited or benchmarked against a public test set.
Voice-clone fraud is rising sharply alongside consumer-accessible voice-cloning tools, making this a durably relevant, fast-growing threat category.
Why CISOs Should Care
Offers a way to verify caller authenticity in real time for high-risk voice channels (collections, KYC, contact centers) without added latency or cloud data exposure.
What Makes It Different
Trains compact models specifically for on-device execution rather than compressing a large cloud model, keeping audio local while still delivering an instant verdict.
The Matrix Verdict
55/100 — INCREMENTAL INNOVATOR
DetectifAI is an early-stage Incremental Innovator: a technically credible, fast-moving founder-led team with real production usage, but funding and independent validation are still minimal.
Editorial Note: Claims vs. Verified Findings
Call-volume and production-use figures come from an independent TechCrunch feature rather than a press release; detection-accuracy specifics were not independently verified.
Sources
Alternatives to DetectifAI
Adaptive Security
AI-driven platform that simulates deepfake, voice, and multichannel social-engineering attacks to train and test organizations against next-generation phishing.
Quilr
Early-stage agentic AI security startup building a 'Service-as-Software' platform to guard against human-related breaches and secure AI agent…
Tenzai
An agentic AI penetration testing startup building autonomous 'AI hackers' to find and validate exploitable vulnerabilities at a…
Zenity
Governance and security platform for AI agents and low-code/no-code development, securing agent identity, permissions and behavior across the…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…
Reco
Reco secures the "agentic ecosystem" — mapping what AI agents can access across SaaS and enterprise apps, detecting…