Spectrum Labs raises $10M for its AI-based platform to combat online toxicity
Every platform with user-generated content faces the same scaling problem: toxic behavior grows faster than human moderators can respond. Keyword filters catch the obvious slurs, but miss the vast majority of harmful interactions — grooming, radicalization, doxxing, and dozens of other patterns that depend on conversational context, not just individual words. At Wing, we backed Spectrum Labs because the company applies AI to this problem at the infrastructure layer, giving platforms a real-time trust and safety system that actually understands what users mean. Today, Spectrum Labs announced a $10 million Series A to expand that system across more languages, more platforms, and more behavior categories. Read our full investment thesis on Spectrum Labs for a deeper look at why this category matters to Wing's AI-first investing lens.
In this post: The platform · The team and the raise · Early traction
What Spectrum Labs built
Spectrum Labs' core product, the Guardian platform, is a dashboard and application programming interface (API) suite that lets platforms integrate contextual AI moderation directly into their workflows. Rather than scanning messages for banned keywords, Guardian analyzes conversation context, metadata, and user history to classify behavior across 40+ toxic behavior profiles — from harassment and hate speech to more nuanced categories like radicalization and grooming. Spectrum Labs developed those profiles in consultation with researchers and academics worldwide.
The platform supports multiple languages and plugs into existing moderation pipelines through APIs. This matters because toxicity is not an English-only problem, and platforms operating globally need detection that works across linguistic and cultural contexts. For Wing, this is the kind of AI-first infrastructure we look for: a technical moat built on proprietary data and models, solving a problem that scales with the internet itself.
The team and the raise
CEO Justin Davis and CTO Josh Newman founded Spectrum Labs after the 2016 U.S. election, motivated by the visible surge in online toxicity and radicalization. Both came from Krux, an online content categorization company for marketers that Salesforce acquired. About nine other Spectrum Labs employees also came from the Krux/Salesforce team. That pedigree matters here: Krux's core competency was categorizing unstructured web content at scale, which is exactly the technical foundation Spectrum Labs needed to classify toxic behavior across billions of online interactions. The team continuity signals deep domain expertise and the kind of founder-team cohesion Wing prioritizes in early-stage investments.
The $10 million Series A was led by Greycroft, with Wing Venture Capital, Ridge Ventures, Global Founders Capital, and Super{set} participating. Total funding reached roughly $14 million. The capital will go toward expanding the platform's language coverage and behavior detection capabilities. For the full details, see the TechCrunch coverage by Ingrid Lunden.
Early traction
What stood out to us about Spectrum Labs' customer base was its breadth across verticals. As of this announcement, the platform serves Riot Games in gaming, Pinterest in social, the Meet Group in dating, and Mercari in marketplace commerce. DTC brands and organizations tracking internal conversations round out the range of use cases. That horizontal spread is meaningful — it confirms that AI-driven content moderation is not a niche feature for social networks but a foundational need for any product where users interact.
Trust and safety infrastructure is one of those categories where the market grows automatically with internet usage — every new platform, every new moderation language, every new behavior profile expands the addressable problem. Spectrum Labs is building the contextual AI layer to match that growth, and Wing is proud to back them.
