Meta Data Scientist Says AI Replaced Her Work Before It Took Her Job

Moyan Chen, a 24-year-old Meta data scientist working on Instagram, was laid off on May 20 in a round that cut about 8,000 Meta employees and reassigned 7,000 more to AI-focused teams. In a first-person account given to Business Insider, Chen said the months of waiting before the cut were hard, and the cut itself felt more like relief than pain.

Chen said the same AI tools her former employer is now racing to build had already absorbed the routine work that defined her role before the company cut it. Her one-line summary of the layoff: “I feel like, ultimately, I lost my job to AI.”

The Layoff Date She Had Been Watching For

Moyan Chen worked as a data scientist on Instagram at Meta, based in New York City. She was laid off in May after just under a year on the job, and her identity and former employment have been verified by Business Insider.

The wait started in March, when rumours of layoffs at Meta leaked internally. Employees had no timeline. “Some of my colleagues and I were fearing Wednesdays because Meta has sometimes laid off people on those days. So, every Tuesday night, when I left work, I wondered if I would come back.” On Wednesday mornings she would wake early and check her email. The pattern lasted about a month, until April, when management gave a date: May 20.

When the date finally arrived, the dread that had been building since March ended. The cut itself, she said, landed as a release.

When the day finally came, and I got laid off, I was like, “This is it.” It was more like relief than pain.

Moyan Chen, a 24-year-old Meta data scientist, in the full first-person account of the May 20 layoff.

The Routine Data Science Work She Says AI Absorbed First

Chen’s verdict on the layoff, in her own words, was simple. “I feel like, ultimately, I lost my job to AI.” She is not the only one inside Meta making that case, but she is unusually direct about which parts of her work the technology had already taken before the layoff date was even set.

Her role at Instagram was a standard data scientist position. The repetitive parts of it, she said, are already being absorbed by AI tools inside the same company that is now spending hundreds of billions of dollars on those tools. The tasks she names are the everyday plumbing of a data scientist at a large social network:

• Writing SQL queries
• Creating data visualizations
• Tracking and analyzing metrics
• Translating business questions into data pulls

“If you only know how to code, that’s not enough,” Chen said. “If you’re just writing SQL queries, using Python, or tracking and analyzing metrics, it’s not a very promising career anymore.” Her summary of the trend is blunt: the title “data scientist” will survive, but the work underneath it will not look the same. “There will still be a role called ‘data scientist,’ but they will need to know more about other functions. There is this emerging trend that requires us to have broader skills and knowledge because of AI.”

The accuracy of the AI tools, in her telling, is what made the human review loop optional. “It got to the point where I wouldn’t check AI-generated queries because they have gotten so accurate. I thought that if AI made a mistake on a specific task, I would make 10. For big, ambiguous projects, AI would still make a lot of mistakes, but for specific tasks, it was super accurate. It’s very much like a talented individual contributor.” That single observation, more than the layoff itself, is the data point that will worry other data scientists reading her account.

The Big Tech Workforce Reshuffle Behind Her Job

Chen is one of about 8,000 Meta employees cut on May 20, in a round that touched roughly 10% of the company’s workforce. Company spokesperson Erica Sackin confirmed to NPR that affected employees had been notified. Headcount at the end of March stood at 77,986, per company filings.

The 8,000 number is only half the trade. Another 7,000 Meta employees were reassigned to new AI-focused teams rather than cut, according to a source familiar with the situation but not authorized to speak publicly, as detailed in the May 20 report on Meta’s 8,000-person layoff. An internal memo from chief people officer Janelle Gale on May 18, reviewed by Bloomberg and Reuters and laid out in the May 18 memo on the 7,000 reassignments to AI teams, named the destination units. The new structure is meant to be “flatter” and built around “smaller teams,” Gale wrote, with employees moving into groups focused on AI products including agents and apps.

By the numbers:

• About 8,000 Meta employees laid off on May 20
• 7,000 Meta employees reassigned to AI-focused teams
• Roughly 10% of Meta’s workforce
• 77,986 Meta employees at the end of March
• More than 1,000 employees signed a petition against the company’s AI training data practices

The money side of the trade is large. In January, Meta forecast capital expenditures this year would be almost double what it spent last year. Meta is also hiring aggressively for AI, luring talent with pay packages that have made compensation data a public topic inside the industry. The Straits Times reported that the new AI units, including Applied AI Engineering and Agent Transformation Accelerator XFN, are aimed at building AI agents that can autonomously carry out tasks currently performed by human staff. The competitive frame, as NPR reported, is unflattering: Meta lags OpenAI, Anthropic, and Google in the AI race. Chen’s individual severance is one slot in a company-wide redirect of cash and headcount into AI, including the AI that already did part of her work.

Why the Big-Company Career Path No Longer Feels Safe

Before the layoff, Chen had run a simple career calculation. She had interned at three large companies, joined Meta, and stayed because the alternative frightened her. “How am I going to feed myself if I don’t work for a big company?” she had asked herself. She did not resign. She kept working and worked hard, in her own words, and for nearly a year the trade had looked like the safest place to be.

The calculation has changed. “Now I feel like it’s not safe anymore, like I can get laid off at any time.” Meta’s severance, by her account, is generous enough to buy her a couple of months to decide what comes next. The old trade, surrender mobility for security, is no longer paying the way it was sold.

Some of her former colleagues are now telling each other that finance is the safer harbour, on the theory that finance adopts AI more slowly. Chen is not convinced. Her image for the moment is the one she used in her account.

It feels like we are all sailing on the sea, and Meta is a huge ship that’s moving very fast. When the AI storm comes, is your next move to jump to a smaller, slower ship?

Moyan Chen, a laid-off Meta data scientist, in her Business Insider account. Her follow-up: “Some people I worked with were saying it’s better to find a job in finance because it takes longer for them to adopt AI. But ultimately, is the same thing going to happen to you?” That is the question she leaves open. Finance firms are asking the same question of themselves, with the answer starting to show in what they actually need from new hires in the AI era.

She is single, has no family in the US, and her parents want her back in China, which she calls the worst-case scenario because she loves the US and the energy of New York City. The layoff is a problem with a runway.

What She Plans to Do Instead

Chen’s bridge from here is concrete. Meta’s severance has bought her a couple of months. She has started creating content online to document her career journey and what she is learning about AI. She is also exploring career coaching for people going through the same transition, a small bet that the experience she just had is one other workers are about to have.

The longer-term bet is directional. She told Business Insider: “I’m less interested in AI as a stand-alone technology and more interested in how it changes the way people work and build products.” If she finds a team whose interests and values line up, she would seriously consider joining an AI startup.

Startups carry risk, and she knows it. “Those companies can be risky, but staying at a big company doing traditional data analytics and reporting jobs just feels like I will be left behind. That’s riskier in the long term.” The data scientist job, in her telling, is being left behind even when the worker stays in it, a point the memo Zuckerberg sent to staff on the AI overhaul also makes, from the top of the company.

Frequently Asked Questions

How many Meta employees were laid off on May 20, 2026?

About 8,000, in a round that touched roughly 10% of Meta’s workforce. The company ended March with 77,986 employees, per company filings, the baseline the 10% figure is calculated against. Meta spokesperson Erica Sackin confirmed the notifications to NPR on May 20.

Did Moyan Chen say AI caused her layoff?

Chen said, “I feel like, ultimately, I lost my job to AI.” That quote comes from the first-person account she gave to Business Insider. It is her stated belief about what happened to her role, reported in her own words rather than as a confirmed finding about the cause of the cut.

Which Meta data science tasks did Chen say AI is already replacing?

Chen’s list of the data science tasks that AI has already absorbed, in her experience at Instagram: writing SQL queries, building visualizations, and tracking metrics. The fourth item she named, translating business questions into data pulls, is the one she said a coding-only data scientist is least equipped to hand off.

What did Meta say it was doing with the employees it didn’t lay off?

Meta reassigned 7,000 employees into new AI-focused teams, according to a May 18 internal memo from chief people officer Janelle Gale, reviewed by Bloomberg and Reuters. The new units, including Applied AI Engineering and Agent Transformation Accelerator XFN, are aimed at building AI products and agents that can carry out tasks currently done by human staff.

Is AI really replacing data scientists, or just changing the role?

Chen argues the title will stay but the work will broaden. The coding-only data scientist, in her telling, is the one being left behind, while the data scientist who can work across business, product, and strategy is the one being asked for.