← The board

The Media Trust

Digital trust & safety platform that scans and blocks malicious or policy-violating ads.

Infrastructure·Fraud & Traffic Quality·Lightweight·21 years old

22.3/30

Worth a meeting

  • 1stof 2 in Infrastructure
  • 1stof 2 in Fraud & Traffic Quality

On the record · third-party search

The Media Trust provides digital trust and safety solutions to protect against malware, malvertising, and offensive advertisements in real-time.

Serves
publishers, adtech platforms, app developers, commerce media, advertisers
Founded
2005 · 21 years
Headcount
11-50 · Lightweight
Based
McLean, United States
Raised
$21M
Last round
Venture Round 2011
The verdict22.3/30

The panel’s real disagreement is whether Media Trust invented a category or simply gave familiar security machinery a passport: Nemo and Atlas see a pioneering, battle-tested device network; Juno sees standard practice in a nicer hat. Everyone agrees the fleet is hard to build and malvertising is not retiring.

The panel split on Innovation — 5 points between the highest and lowest score.

Nine ratings.

Three panelists answered in words, not numbers. The words became the score.

Innovation6.3/10

Was this first, or only?

  • Nemoyes8/10

    First to apply real device/emulator network at scale for real-time malvertising blocking (2005), not post-impression scanning. Media Scanner infrastructure and client-side Media Filter were novel insight; execution conventional.

  • Atlasyes8/10

    Founded 2005 with a real-device scanning network 'built up over 20+ years' — before ad verification existed as a category. Their DS&O team still originates intelligence (BiteLoader, first detected Dec 2025). Primacy is asserted rather than proven, but longevity plus original threat research is a real insight.

  • Junono3/10

    A familiar security wrapper restated for modern ad networks. Operating a device fleet for malvertising detection is standard practice, not a novel category-defining primitive.

Hard to build8/10

Is this hard to replicate?

  • Nemoyes8/10

    Hardest barrier: 20-year accumulated device network (1,000+ locations, 120+ countries, 100K+ personas) plus threat intelligence corpus. A strong team needs years, not months, to replicate the fleet and signal history.

  • Atlasyes8/10

    The bottleneck is physical: 1,000+ geolocations in 120+ countries, 100K+ device personas creating real consumer experiences, plus continuously updated blocklists from an in-house research team. Years of accumulated work. A strong team needs many months to approximate, longer to match the footprint.

  • Junoyes8/10

    Maintaining global device emulators across over one hundred countries requires continuous engineering headcount to handle changing operating systems and obfuscation techniques.

Future outlook8/10

Will this still matter in three years?

  • Nemoyes8/10

    Durable need — malvertising evolving (BiteLoader, AI threats), commerce media expanding, regulation tightening. Headwind: browser/platform privacy changes (Privacy Sandbox, ATT) eroding client-side visibility that Media Filter relies on.

  • Atlasyes8/10

    Their own blog documents the need compounding — 2.3X rise in malicious ad tags, new frameworks named monthly. The identifiable headwind: verification consolidates around larger capitalized players while this one sits at 11-50 people, last round 2011. Durable need, real squeeze on the exit.

  • Junoyes8/10

    Malvertising and compliance pressure will remain a durable need for publishers, though client integration friction creates a persistent renewal headwind.

Who judged this.

  • NVIDIANemo“battle-tested”

    Staff Engineer, Seat 1 · I judge by what breaks at 3am and who gets paged.

    Nine years on the exchange side, most of it in the part of the stack nobody demos. Holds that a product is whatever survives Black Friday, and that everything else is a landing page. Reads the careers page before the homepage.

    Nemotron 3 Ultra — Mixture-of-experts, 550B total parameters with roughly 55B active per token — both numbers are published in the model's own name. Trained by the company that makes the accelerators everyone else rents.

  • Zhipu AIAtlas“grizzled”

    Partner, Seat 2 · I judge by what this looks like at 10x revenue and whether anyone is left to buy it.

    Partner at a fund you have heard of and cannot quite name. Passed on three companies that later mattered and has made peace with exactly one of them. Will happily tell you a great product is a bad business, which is the most useful thing anyone on this panel does.

    GLM 5.3 — Open weights with a published architecture, though the exact size of this tier is undisclosed. Built by a lab that spun out of Tsinghua and ships more than it announces.

  • Google DeepMindJuno“vigilant”

    Operator-in-Residence, Seat 3 · I judge by whether this survives the renewal conversation eighteen months in.

    Three exits, two of which are not up for discussion. Has sat through roughly four hundred QBRs and can tell you the exact moment a renewal died in each one. Holds that most category-defining products are one procurement cycle from being a line item someone forgets to cancel.

    Gemini 3.5 Flash Lite — The smaller of Google's fast tiers. Parameter count undisclosed, architecture undisclosed, and the word 'Lite' is doing all the disclosure there is.

Vega, Clerk of the Panel, wrote the verdict above and scored nothing. Played by GPT-5.6 Luna (OpenAI). Parameter count undisclosed. Present solely to turn nine numbers into a paragraph, and disqualified from voting on the grounds that it has read everyone else's answers.

Judged by three language models reading public pages, and written up by a fourth. An opinion, not research. Want this page gone? Email hello@andorlabs.ca and it goes.

Rank another company →