About FrameBright.
FrameBright.ai is the technology and licensing surface for the AEGIS classification engine. We build the inference layer for safety-critical and high-stakes content environments.
What we do
FrameBright builds AEGIS, a content classification engine purpose-built for environments where adversaries actively engineer content to defeat filters. The engine is licensed to ISPs, device OEMs, network-equipment vendors, and enterprise customers as a network-edge or platform-embedded component, not rented as a third-party API.
The .ai surface is the platform-licensing and partner relationship surface. Consumer users of the FrameBright brand visit framebright.com. Community-driven content classification lives at parentproof.com and parentproof.org. The classification engine that runs underneath all of these is the same: AEGIS.
Why we exist
The classification of content on the open internet has historically been built one of three ways. Network-layer DNS blocklists, which catch nothing dynamic and produce more false positives than true ones. Application-layer LLM moderation, which is brittle against adversarial content engineered to defeat it and expensive at scale. Closed proprietary classifiers operated by hyperscale platforms, which work well for the platforms that own them but are unavailable to the rest of the ecosystem.
The result has been an inference gap. Telecoms cannot offer parental and safety controls that hold up against modern content. Device manufacturers cannot embed classification at the hardware boundary. Independent platforms cannot afford either the model investment or the moderator headcount required to run safety in-house.
FrameBright closes that gap. AEGIS is the classification engine that platform partners license, embed, and run themselves.
Where we come from
The founders operated content-safety systems at consumer-internet scale before founding FrameBright. CTO Eric MacDougall ran platform engineering for one of the highest-volume adversarial content surfaces ever built (45M+ monthly active users), then architected real-time game servers at 200K+ concurrent for partnerships with EA, then ran video-intelligence engineering for commercial partnerships across Google and AWS, then built agent-to-agent payment protocols alongside Google's A2A team, Mastercard, and Visa. The classification choices in AEGIS draw directly from that operating history; this is not classification theory.
FrameBright is part of Good Ventures Lab, the venture studio building coordinated AI infrastructure across nine products: R1 (agent runtime), RelayGate (router-core), RelayOne (governance), TrueCom (signed receipts), Actium (managed AI workforce), Heroa (sovereign substrate), DeepTap (private-corpus search), Veritize (drift-tracking), and FrameBright (classification). Each piece works on its own; together they compose into a coordinated trust stack for partners deploying AI in production.
How we are commercially different
AEGIS is licensed, not sold by the seat. Partners embed AEGIS as a component in their own products. They own their integration, own their latency, own their customer relationship. This is closer to how classification infrastructure has historically been procured by major telcos and hardware OEMs than to the modern SaaS moderation API model. Partners who only want AEGIS get AEGIS; the connected stack is available, never required.
Four licensing paths match the four deployment surfaces: ISP CPE, device-embedded, router/network, and managed-API enterprise. Each has its own commercial shape, integration effort, and performance contract. Detail at /licensing/.
Get in touch
Licensing inquiries: contact form or [email protected]. Press: press page. Research collaboration: research surfaces.