Scale AI
NativeAI NativePrivateUpdated November 15, 2026The company that made AI training possible for every major lab — and is now building its own intelligence layer.
- Website
- scale.com
- Founded
- 2016
- CEO
- Francis deSouza
- HQ
- San Francisco, CA, United States
- Stage
- Private
- Revenue
- $500M–$1B
- Total Funding
- $15.9B
- Employees
- 501-1000
Executive Summary
Updated Nov 12, 2026
What they do
- Supplies human-labeled and expert training data for frontier AI model developers
- Sells GenAI and agent platforms that help enterprises and governments build AI
- Runs AI model testing and leaderboards through its SEAL research lab
Why it matters
- Pentagon AI office raised Scale's agreement ceiling to $500 million in May 2026
- Reported revenue of $870 million in 2024, guiding to about $1 billion for 2026
- Meta's $14.3 billion stake in 2025 valued the company at roughly $29 billion
Best fit
- Defense and public-sector agencies needing cleared, mission-focused AI deployment partners
- Large enterprises wanting help taking generative AI from pilot into production
- Model builders needing large-scale data labeling, RLHF and evaluation services
Watch-outs
- Meta's 49% stake raises data-neutrality concerns for labs that compete with Meta
- Several major AI labs reportedly moved data work to rivals after the Meta deal
- Leadership changed in 2025, and Droege has been described as interim CEO
Products & Offerings
Scale Data Engine
Scale's main data labeling, RLHF and expert data platform for training and fine-tuning AI models.
Scale GenAI Platform
Enterprise platform for building, fine-tuning, evaluating and deploying generative AI apps and agents on a company's own data.
Scale Donovan
AI decision-support and agent platform for defense and national-security users, used in Scale's Pentagon work.
Scale Evaluation and SEAL Leaderboards
Model testing, red-teaming and public leaderboards for measuring how well large language models perform and how safe they are.
Outlier
Platform that recruits and manages freelance experts who produce training data for AI models.
Latest News
Unite.AI · 3 days ago
PositiveScale AI · 3 days ago
PositiveGulf Times · 6 days ago
PositiveFourWeekMBA · 10 days ago
PositiveSt. Louis Call Newspapers · 13 days ago
PositiveUpdated Nov 12, 2026
Scale AI is a large, proven provider of AI training data, model testing and government AI work. Its Pentagon deal is growing and its revenue is near $1 billion, so it belongs in any structured evaluation in this category. Lower 2026 revenue guidance, customers leaving after the Meta deal, and an interim-era leadership team mean buyers should check its neutrality, data-isolation controls and staffing before committing. Defense and enterprise buyers who don't compete with Meta face the least conflict.
Signal Provenance
Latest observation date and source for each signal category
| Signal Category | Last Updated | Source |
|---|---|---|
Funding Signals Last funding round, Funding stage | Nov 15, 2026 | Cited web research |
Hiring Velocity Headcount, Headcount growth | Nov 13, 2026 | Research agents (estimate) |
Technical Output Research papers, GitHub activity | Nov 15, 2026 | Hugging Face Hub + news feed + research agents |
News Sentiment News mentions, News sentiment | Nov 15, 2026 | Verified news feed |
Partnership Signals Enterprise partnerships, Distribution partners | Nov 15, 2026 | Research agents (estimate) |
Revenue Indicators ARR estimate, Revenue growth | Nov 15, 2026 | Research agents (estimate) |
Leadership Signals CEO tenure, Founder-led | Nov 15, 2026 | Research agents (estimate) |
Competitive Position Category rank, Market share | Nov 15, 2026 | Research agents (estimate) |
Why this Bluescore?
6 dimensions, each scored 0–10 against all 250 companies. Each adds its weight × score to the Bluescore — see methodology
Scale AI's Bluescore of 65.1 rests most on Investment Momentum (8.4, top 5%) and Talent & Research (8.6, top 7%). Its weakest area is Product Depth (4.5, 41st percentile).
In plain English
Scale AI scores 65.1 out of 100 and ranks #19 of 247, lifted most by Talent & Research at 8.6 of 10 (93rd percentile among the AI250) and Investment Momentum at 8.4 of 10 (95th percentile among the AI250). Product Depth holds it back: it scores 4.5 of 10 (41st percentile among the AI250) and leaves 11.6 of its 21 weighted points unfilled, the largest gap of any dimension. Improving Product Depth would move the Bluescore more than any other change, because its 21% weight is among the heaviest and its gap is the widest.
Written by AI from the scores below and checked against them; updated nightly.
Risks
Updated Nov 12, 2026