Modulate
Modulate builds voice AI models that detect fraud, deepfakes, and toxic behavior in real time across gaming, contact center, and voice-agent platforms.
Visit Website ↗ + Add to Compare Claim This CompanyOverview
Modulate develops voice AI technology originally built for real-time content moderation in online games (its ToxMod product), and has since expanded into fraud detection, deepfake/voice-clone identification, and compliance monitoring for AI voice agents used in regulated industries such as finance and healthcare. The platform runs more than 100 specialized small models to transcribe calls, detect emotion and intent, flag synthetic voices, and enforce compliance rules without requiring humans to review every interaction.
Founded in 2017 in Somerville, Massachusetts by Mike Pappas and Carter Huffman, Modulate has raised roughly $66 million to date, including a $25 million round in September 2026 led by Future Ventures with Hyperplane and Lakestar participating. The company’s shift from gaming trust-and-safety into enterprise fraud and deepfake detection reflects the broader convergence of voice-based social engineering risk with AI security.
Innovation Matrix Assessment
The company has successfully pivoted its core voice-AI technology from gaming moderation into fraud, deepfake, and compliance use cases across multiple funding cycles since 2017.
Automating voice-channel risk monitoring (fraud, deepfakes, compliance) reduces the manual review burden on trust-and-safety and fraud teams handling high call volumes.
Roughly $66M raised to date from named investors (Future Ventures, Lakestar, Hyperplane), reported independently by TechCrunch, is solid, verifiable momentum for the category.
Real-time multi-model voice analysis is a genuine technical advance, but it extends and consolidates existing trust-and-safety and fraud-detection categories rather than replacing them outright.
Detection-accuracy and scale claims are vendor-reported; no independently audited efficacy study or named-incident case study was found in this research.
Voice-based fraud and deepfake scams are increasing quickly as voice-cloning tools become more accessible, keeping this category durably relevant.
Why CISOs Should Care
Gives security and fraud teams automated coverage of voice channels — calls, IVR, voice agents — that are otherwise hard to monitor at scale for social engineering and deepfake attacks.
What Makes It Different
Runs a large ensemble of purpose-built small models per call rather than one general classifier, enabling simultaneous fraud, deepfake, emotion, and compliance signals in real time.
The Matrix Verdict
65/100 — INCREMENTAL INNOVATOR
Modulate is an Incremental Innovator approaching Meaningful: strong funding and a credible pivot into enterprise fraud/deepfake detection, held back by a lack of independently verified efficacy data.
Editorial Note: Claims vs. Verified Findings
Funding and founding facts are independently reported by TechCrunch and CB Insights; specific detection-performance claims are vendor-stated and not independently verified here.
Sources
Alternatives to Modulate
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…
Charm Security
Agentic AI workforce that investigates and intervenes on scams and fraud in real time, reading manipulation and intent…
Alice (formerly ActiveFence)
Israeli AI security company (rebranded from ActiveFence in January 2026) offering a lifecycle platform to test, guard, and…