Why Netflix ties AI freedom to strict ownership

This article summarizes an episode of Lenny’s Podcast’s video series featuring Elizabeth Stone, chief product and technology officer at Netflix.

Elizabeth Stone, chief product and technology officer at Netflix / Photo credit: Netflix

AI is making it easier for employees to turn an idea into something that looks real. However, this also makes it easier for poorly thought-out ideas to reach customers much faster.

Elizabeth Stone, the chief product and technology officer at Netflix, views this as a major challenge for managers. AI allows more employees to create new products. It also forces companies to decide who is responsible for what actually gets released to customers.

When more people can build, ownership has to get clearer

At Netflix, AI speeds up early work by letting employees who are not engineers test their ideas before the engineering team gets involved. Still, complete-looking test versions can make a proper review feel unnecessary.

In reality, deep expertise is still needed to decide if a new feature is truly ready to be released. Stone sets a strict limit when it comes to the final product. At this final stage, being fast is less important than being responsible.

She explains, “Anytime a new technology comes along, especially one as transformative as GenAI, you go through a storming phase before forming. That does not mean we should put AI back into the box… Do I believe anyone [using AI] should be shipping code to production? Probably not.”

This limit allows Netflix to encourage employees to create test models while ensuring that technical experts remain responsible for what reaches customers. Wide experimentation is welcome, but someone specific must take ownership of the final released product.

Understanding AI becomes a way to evaluate talent

Because AI allows more people to help create early product ideas, companies are changing how they evaluate their staff, focusing on how well employees use judgment to assess their own work.

“AI fluency is going to vary by function and based on where you are in your career,” Stone explains. “It means using technology where it’s useful, exercising good judgment and having the mindset to explore and try new things. That’s non-negotiable for all roles.”

This expectation means that basic AI skills are required for hiring and job performance, even for company leaders who never write computer code. Human resources teams must evaluate an employee’s judgment in real situations. The goal is to reward a true understanding of the technology.

Stone’s approach also protects the hiring of junior staff. Netflix still hires interns and recent graduates because they naturally understand modern technology and current consumer trends.

In exchange, the company provides these younger employees with more guidance on maintaining quality and reviewing their work. This training ensures that quick computer-generated results are backed up by a deep understanding of the job.

Big picture thinking is a crucial advantage when using AI

A simple shortcut made by one person using AI can create a risk for the entire company. Stone solves this by building shared company-wide systems.

High standards limit how many new rules Netflix adds