What is Scale AI?
Data, evaluation, and infrastructure platform for building, fine-tuning, and deploying AI systems.
- Category
- AI data and evaluation
- Headquarters
- San Francisco, CA
- Founded
- 2016
- Employees
- ~1,000
- Total funding
- $15B+ incl. Meta
- Valuation
- ~$29B (2025 Meta deal)
What is Scale AI?
Scale AI provides data labeling, data generation, model evaluation, RLHF, safety, and government AI infrastructure for frontier labs, enterprises, and defense customers.
Scale AI started as a data-labeling company for autonomous vehicles and computer vision, then expanded into large-language-model data, reinforcement learning from human feedback, model evaluation, safety, and public-sector AI programs. Its customers have included leading AI labs, enterprises, and U.S. government agencies.
In 2025 Meta made a $14.3 billion strategic investment for a large minority stake, valuing Scale around $29 billion and bringing founder Alexandr Wang into Meta's AI leadership. That deal changed Scale's leadership and customer dynamics while confirming the strategic value of high-quality training and evaluation data.
What does Scale AI offer?
Scale offers data labeling, GenAI data, model evaluation, RLHF, synthetic data, safety, Donovan for government, and enterprise AI applications.
- Scale Data Engine· Data
- Generative AI Data Engine· LLM data
- Model evaluation· Evaluation
- RLHF· Alignment
- Synthetic data· Data generation
- Scale Donovan· Government AI
- Safety and red teaming· AI safety
Sources:Scale productsScale government
How does Scale AI make money?
Scale makes money from enterprise and government contracts for data labeling, expert data, model evaluation, RLHF, AI applications, and managed services.
Scale's pricing is custom rather than public self-serve pricing. Contracts depend on data type, expert labor requirements, quality thresholds, security controls, volume, annotation complexity, evaluation design, and whether the customer is commercial, frontier-lab, or government.
Growth is driven by AI labs needing higher-quality data and evaluations, enterprises deploying models into production, and government agencies buying AI readiness and mission systems. The Meta investment also creates strategic concentration risk because some AI lab customers may reconsider vendor neutrality.
Sources:Scale productsScale government
Who leads Scale AI?
Scale AI's post-Meta leadership centers on CEO Jason Droege, while founder Alexandr Wang moved to Meta and remains central to Scale's history.
- Jason DroegeChief Executive OfficerCEO after 2025 Meta transactionFormer Uber Eats leader brought operating experience to Scale's next phase.
- Alexandr WangFounder and former CEOFounder - 2016; moved to Meta in 2025Founded Scale and became a key leader in Meta's AI effort after the strategic investment.
- Lucy GuoCo-founderCo-founder; departed earlierCo-founded Scale and later founded Passes.
- Michael KratsiosManaging Director / Public sector advisorPublic-sector leadershipPublic-sector and national-security relationships are strategically important to Scale.
How do you contact Scale AI's leadership?
Scale publishes sales, government, and press channels; verified personal executive emails are not public. Executive addresses below are format-following only.
first.last@scale.com (format-following; verify before use)Sources:Scale contactScale newsroom
How much funding has Scale AI raised?
Scale AI has raised more than $15 billion including Meta's $14.3 billion strategic investment, which valued the company around $29 billion in 2025.
Scale's major rounds include early seed and Series A/B funding after its 2016 founding; a $100 million Series C in 2019 at a unicorn valuation; a $155 million Series D in 2020 at about $3.5 billion; a $325 million Series E in 2021 at about $7.3 billion; and a $1 billion Series F in 2024 at about $13.8 billion.
The defining financing event was Meta's 2025 investment of about $14.3 billion for a 49% stake, valuing Scale around $29 billion and bringing Alexandr Wang to Meta. That transaction was more strategic than a normal venture round because it tied Scale's data and evaluation capabilities to Meta's frontier AI ambitions.
How did Scale AI get here?
Scale moved from autonomous-vehicle labeling into core infrastructure for generative AI and defense AI.
- 2016FoundedAlexandr Wang and Lucy Guo founded Scale AI.
- 2019Series CRaised $100M and reached unicorn status.
- 2021Series ERaised $325M at about a $7.3B valuation.
- 2024Series FRaised $1B at about a $13.8B valuation.
- 2025Meta strategic investmentMeta invested $14.3B for a large minority stake at a ~$29B valuation.
- 2025-2026Post-founder CEO phaseLeadership shifted after Alexandr Wang moved to Meta.
Who are Scale AI's competitors?
Scale competes with AI data-labeling platforms, model-evaluation vendors, synthetic-data companies, and in-house AI lab data teams.
- TuringAI data, coding, and talent platform for training and evaluating models.
- LabelboxData labeling and model-evaluation platform with enterprise data workflows.
- AppenPublic data-annotation and AI-training-data provider with global crowd operations.
- Surge AIRLHF and high-quality data provider focused on frontier AI labs.
- Invisible TechnologiesAI process automation and human-in-the-loop operations provider.
Sources:Scale productsLabelbox
Scale AI — frequently asked questions
