Public transcript

Alexandr Wang, Scale AI CEO

By: E. Valley
Length: 5 min
Language: English
Text from: the video's captions
Transcribed: 12 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

Alexander Wang, founder and CEO of Scale AI, delivers testimony to a congressional committee on ensuring US leadership in AI. He outlines China's four-pronged AI strategy and argues the US must chart its own course anchored in American values. He proposes four themes for action: dominate on data, unleash AI in government, innovate through smart regulation, and promote US technology globally.

Key points

Alexander Wang grew up in Los Alamos, where his parents were physicists at the National Lab, the birthplace of the atomic bomb.
He explains that progress in AI depends on three key elements: data, compute, and algorithms.
China's AI plan includes a whole-of-country approach, outinvesting the US in data, catching up on compute, and exporting leading AI models.
Wang proposes establishing a national AI data reserve as a centralized hub for government AI work.
He calls for an 'agentic government' that uses AI under human supervision, with each agency setting up at least one flagship program.
He recommends a use case-based regulatory framework and a single federal AI governance standard to avoid state-level patchwork legislation.
He urges empowering NIST to complete AI measurement science and export it through the global network of AI safety institutes.
He describes countries that will choose between Western and CCP-controlled technology as 'AI geopolitical swing states.'

Questions it answers

00:00Witness background

Who is Alexander Wang and why is he testifying?

He is the founder and CEO of Scale AI, which provides expert-level data and technology solutions to AI labs, enterprises, and the US government. He grew up in Los Alamos with physicist parents at the National Lab, which shaped his belief that science and technology leadership is vital to national security.

AI progress depends on data, compute, and algorithms.
Wang started Scale because few at MIT focused on the data challenge.
Scale keeps humans at the center of everything it does.

01:00China's AI plan

What is China doing in AI?

The CCP's AI master plan, taking shape since 2018, now includes a whole-of-country approach via the AI plus initiative, outinvesting the US in data, catching up on compute and infrastructure, and developing and exporting leading AI models.

China is spending billions on AI-ready data and unlocking public data sets.
China uses AI to surveil and suppress its people.
The competition is close across all foundational elements.

02:09Dominate on data

How can the US win on data?

Wang proposes establishing a national AI data reserve as a centralized hub, making all government data AI ready with proper infrastructure, and having Congress invest to make data dominance a national priority.

The US government is one of the largest data producers in the world but most data is unavailable for AI.
The data reserve would enable sharing between agencies.

03:21Unleash an agentic government

How should the government use AI?

Wang calls for an agentic government that uses AI under human supervision, such as streamlining veteran healthcare paperwork, improving IRS fraud detection, and boosting inter-agency information sharing. Congress should require each agency to set up at least one flagship agentic government program.

AI use should improve the lives of public employees and the American people.
Human supervision remains central to the vision.

04:27Innovate with smart regulation

What regulatory framework does Wang recommend?

He recommends a use case-based regulatory framework with an analysis of regulatory gaps, a single federal AI governance standard to avoid state-level patchwork legislation, and policies that train American workers to become the AI workforce of the future.

The framework should allow innovation while creating proper guardrails.
Workers need skills to train, fine-tune, and evaluate AI systems.

Notes

Background and Witness Introduction

Alexander Wang is the founder and CEO of Scale AI, which delivers expert-level data and technology solutions to AI labs, multinational enterprises, the US government, and allies.
He grew up in Los Alamos, where his parents were physicists at the National Lab, the birthplace of the atomic bomb.
At MIT, he learned that AI progress depends on three key elements: data, compute, and algorithms. Few classmates focused on data, which inspired him to start Scale.
Scale keeps humans at the center of everything, because AI should work for us, not the other way around.

China's AI Strategy

The Chinese Communist Party's AI master plan began taking shape in 2018, using the technology to surveil and suppress their people.
The plan now has four focus areas:
A whole-of-country approach, including the recently launched AI plus initiative.
Outinvesting the US in data, spending billions on AI-ready data and unlocking public data sets.
Catching up on compute and building physical infrastructure.
Developing leading AI models and exporting them to the world.

The US Response: Four Themes

Dominate — win on data:
Establish a national AI data reserve as a centralized hub for government AI work, enabling data sharing between agencies.
Make all government data AI ready and stand up AI data infrastructure.
Congress should invest to position data dominance as a national priority.
Unleash — establish an agentic government using AI under human supervision:
Examples: streamlining veteran healthcare paperwork, improving IRS fraud detection, boosting efficiency and information sharing.
Congress should require each agency to set up at least one flagship agentic government program.
Innovate — maximize companies' ability to innovate:
Confirm a use case-based regulatory framework and analyze regulatory gaps.
Establish a single federal AI governance standard to avoid state-level patchwork legislation.
Implement policies enabling American workers to train, fine-tune, and evaluate AI systems.
Promote — promote US technology globally:
'AI geopolitical swing states' will choose between Western or CCP-controlled technology.
Empower NIST to complete AI measurement science and export it through the global network of AI safety institutes.