Meta Deepens AI Push With Key Hire From OpenAI

Key Takeaways

• Trapit Bansal, former OpenAI researcher, has joined Meta to enhance its reasoning AI models.

• Meta now boasts multiple ex-OpenAI researchers on its team, signaling a growing talent shift.

• The company acquired a 49% stake in Scale AI, valued at nearly $15 billion.

• Meta signed a 20-year deal for 1.1 GW of nuclear power to fuel AI data centers.

Former OpenAI Researcher Joins Meta’s AI Team

Bansal played a central role in the development of OpenAI’s o1 reasoning model, a critical project aimed at advancing machine reasoning. His expertise further solidifies Meta’s growing portfolio of high-profile AI experts.

A Growing Roster Of AI Talent

Bansal joins a growing list of ex-OpenAI researchers now working at Meta, including:

• Lucas Beyer

• Alexander Kolesnikov

• Xiaohua Zhai

This brainpower consolidation reflects Meta’s aggressive efforts to dominate the AI space and signals a shift in where top AI talent is heading.

Real-World Intelligence For AI

Meta’s strategy revolves around making AI systems more capable by grounding them in the physical world. The company’s Chief AI Scientist, Yann LeCun, emphasized the company’s core AI objectives:

“There are four essential characteristics of intelligent behavior that every animal, or relatively smart animal, can do, and certainly humans: understanding the physical world, having persistent memory, being able to reason, and being able to plan complex actions, particularly planning hierarchically.”

Meta Will Utilize Video Data & Robotics To Better Train AI Models

Source: Meta

To achieve this, Meta is investing in training AI on real-world data, from video streams to robotics, helping machines learn in ways that mirror human development.

A Year of Strategic AI Expansion

2025 has been a landmark year for Meta in terms of AI-related investments, partnerships, and infrastructure development. Let’s take a closer look.

In June 2025, Meta purchased a 49% stake in Scale AI, a data-labeling powerhouse used across the AI industry. The company is valued at nearly $15 billion.

As part of the deal, Scale AI CEO Alexandr Wang has joined forces with Meta’s internal teams to collaborate on AI model improvements and data labeling efficiencies.

Powering AI Ambitions With Nuclear Energy

To support its growing data infrastructure, Meta signed a 20-year clean energy deal with Constellation Energy, securing 1.1 gigawatts of nuclear power.

The electricity will be sourced from Constellation’s Clinton Clean Energy Center in Illinois, starting in 2027. This energy will power Meta’s AI-dedicated data centers, aligning with its sustainability goals.

Military-Tech Partnership With Anduril

In May 2025, Meta announced a partnership with Anduril, a defense technology company specializing in AI and autonomous systems.

The partnership aims to produce AI-powered AR headsets for the U.S. military, integrating Anduril’s Lattice AI platform—a real-time system that compiles battlefield data from multiple sources.

Why This All Matters

These strategic hires and deals reflect Meta’s clear intent: to become a global leader in the AI arms race. As governments and corporations around the world ramp up investments, Meta is ensuring it doesn’t fall behind.

Mark Zuckerberg Has Repeatedly Stated He Wants Meta To Be A Global AI Leader

Source: Techcrunch

By training AI with real-world data, investing in clean energy, and strengthening ties with defense and data labeling companies, Meta is setting the foundation for long-term dominance.

FAQ

Who is Trapit Bansal?

Trapit Bansal is a former lead researcher at OpenAI, known for his work on the o1 reasoning model. He recently joined Meta’s AI division.

What is the o1 reasoning model?

The o1 model is a reasoning AI system developed by OpenAI to enhance logical processing and complex decision-making in machines.

Why did Meta invest in Scale AI?

Scale AI provides essential data labeling services that are critical for training large AI models. The investment allows Meta to optimize data preparation for its own systems.