Spectrum Labs Raises $32M for AI Content Moderation

Combating Online Toxicity: Spectrum Labs Secures $32 Million in Funding

As online interactions continue to dominate daily life, the prevalence of toxic exchanges remains a significant concern. Many individuals are experiencing negative encounters within these digital spaces. Addressing this issue, Spectrum Labs, a company specializing in artificial intelligence for real-time toxicity detection, has announced a $32 million funding round.

Investment Details and Future Plans

The funding will be allocated to further development of the company’s technology, expansion of its consumer-facing services, and a new venture into enterprise solutions. Spectrum Labs aims to provide businesses with tools to monitor both internal and external communications, offering not only toxicity detection but also comprehensive audit trails for trust and safety initiatives.

CEO Justin Davis articulated the company’s vision: “We aspire to be the leaders in language where civility matters.”

Intel Capital spearheaded the investment round, with participation from Munich Re Ventures, Gaingels, OurCrowd, Harris Barton, and existing investors including Wing Venture Capital, Greycroft, Ridge Ventures, Super{set}, and Global Founders Capital. Greycroft previously led Spectrum’s $10 million Series A round in September 2020, bringing the company’s total funding to $46 million.

Platform Integration and Technological Capabilities

While the company is not disclosing its current valuation, co-founder Josh Newman, the CTO, indicated that business performance is strong. Spectrum Labs currently collaborates with over 20 major platforms, including Pinterest, The Meet Group, Grindr, Fandom, Riot Games, and Udemy. These platforms collectively serve millions of users who exchange billions of messages.

The core of Spectrum Labs’ technology lies in natural language processing, operating in real-time on both text and audio interactions. Notably, the audio analysis doesn’t rely on transcription, enabling faster responses to potentially harmful content.

Davis highlighted the advantage of this approach, contrasting it with the typical delay associated with audio monitoring: “a platform has to wait for users to flag iffy content, then the platform has to find that audio in the transcriptions, and then it can take action — a process that could take days.”

Expanding Behavioral Profiles and Language Support

Spectrum Labs scans communications for over 40 distinct behavior profiles, developed in collaboration with researchers and academics. These profiles encompass harassment, hate speech, violent extremism, scams, grooming, illegal solicitation, and doxxing. The system currently supports nearly 40 languages, with the capability to expand to any language within weeks.

“We can technically cover any language in a matter of weeks,” Davis stated.

Balancing Free Speech and User Safety

The company recognizes the delicate balance between free speech and online safety. A key focus will be empowering users with greater control over their online experience, alongside robust tools for platform trust and safety teams.

Davis acknowledged the inherent tension: “There is a natural tension between what the policy implements and what users want and are willing to accept.” He emphasized the platform’s role in “keeping the worst of the worst off, but also to provide consumer controls to make selection over what they want to see over time.”

Enterprise Solutions and Competitive Landscape

In addition to consumer-focused services, Spectrum Labs is expanding into the enterprise market. This includes monitoring internal communications, as well as interactions with customers in areas like sales, customer service, and marketing. The gathered data can then be used to optimize these operations.

Key competitors in the enterprise space include Aware, which focuses exclusively on enterprise conversation monitoring. L1ght is another competitor specializing in the consumer market.

Strategic Partnership with Intel

Intel is joining this funding round as a strategic investor. The collaboration will involve integrating Spectrum Labs’ technology with Intel’s chip designs, enhancing processing speed and offering a unique selling point for Intel’s hardware customers prioritizing trust and safety.

Mark Rostick, VP and senior MD at Intel Capital, commented: “We believe Spectrum Labs’ Natural Language Understanding technology has the potential to become the core platform that powers the trust initiatives of thousands of companies around the world.” He further stated, “As digital trust and ethical operations emerge as a key factor to help organizations differentiate themselves, we see a huge opportunity to build a Trust & Safety tech layer into enterprise operations.”