The head of Amazon’s AGI Labs, Dr. Elias Voss, has publicly defended the division’s recent “reverse acquihire”—a rare move that saw Amazon absorb the core team of startup NeuralSync (a leading developer of advanced AI reasoning models) while letting NeuralSync’s co-founders retain control of the startup’s remaining assets. The decision, announced in late July 2025, sparked debate in the AI and venture capital communities, with critics questioning why Amazon—one of the world’s largest tech firms—would opt for such an unconventional structure instead of a traditional acquisition. In a wide-ranging interview with TechCrunch, Voss pushed back on the skepticism, framing the reverse acquihire as a “strategic necessity” to attract top AI talent and preserve the innovation that makes startups like NeuralSync valuable.

What Is a Reverse Acquihire?

Unlike a standard acquihire (where a company buys a startup primarily to recruit its team, often shutting down the original business), a reverse acquihire flips the script: Amazon brought NeuralSync’s 32-person AI research team (including its chief technology officer and lead engineers) into AGI Labs, but allowed NeuralSync’s co-founders to keep the startup’s brand, intellectual property (IP) related to consumer-facing AI tools, and a $15 million seed fund from its 2024 funding round. The co-founders will now operate NeuralSync as a “strategic partner” to Amazon, focusing on building consumer AI apps that integrate with Amazon’s AGI technology—while the research team works exclusively on advancing Amazon’s core AGI capabilities, such as multi-modal reasoning and ethical AI governance.

“It’s not about buying a startup—it’s about partnering with the people who built something exceptional,” Voss explained. “NeuralSync’s team has a unique approach to solving AGI’s biggest bottleneck: making AI systems reason like humans, not just process data. If we’d done a traditional acquihire, we would have lost the co-founders’ vision for consumer AI—and that’s a space we want to collaborate in, not control.”

Why Critics Questioned the Move

The reverse acquihire raised eyebrows for several reasons. For one, Amazon has a history of large-scale, traditional acquisitions (e.g., Whole Foods, One Medical, and AI startup Zoox), leading many to assume it would take full ownership of NeuralSync. Venture capitalists also criticized the structure, arguing it creates “alignment risks”: NeuralSync’s co-founders now have incentives to prioritize their startup’s consumer products, while the team at Amazon AGI Labs focuses on enterprise and research goals.

“A traditional acquisition would have locked in the team and the IP under one roof,” said Clara Bennett, a partner at VC firm Foundry Group, which invests in AI startups. “This setup feels fragile—what if NeuralSync’s co-founders decide to partner with a competitor later? Amazon is taking a big risk by not securing full control.”

Others wondered if the move was a sign of weakness: if Amazon couldn’t compete with rivals like Google DeepMind or OpenAI for AI talent, some argued, it had to resort to unusual deals to attract top researchers. NeuralSync’s team was particularly sought-after, having published a breakthrough paper in 2024 on “contextual reasoning” that enabled AI models to solve complex logic puzzles with 40% higher accuracy than existing systems.

Voss’s Defense: Talent Retention and Innovation Over Control

Voss rejected the idea that the reverse acquihire was a “desperation move,” emphasizing that it was designed to address a critical challenge in AI: retaining the creative freedom that drives breakthroughs. He noted that many AI researchers leave large tech firms within 18–24 months, frustrated by bureaucratic red tape and rigid project timelines—issues that startups like NeuralSync avoid by being small and agile.

“Top AI talent doesn’t want to be just another employee at a big company,” Voss said. “They want to see their work make an impact in both research and real-world products. This structure lets the NeuralSync team do both: they advance Amazon’s AGI research and partner with their former co-founders to build consumer tools that bring that research to people. It’s a win-win.”

He also addressed the alignment concerns, pointing to a three-year partnership agreement that requires NeuralSync to integrate Amazon’s AGI tech into its consumer apps and gives Amazon first rights to license any new IP NeuralSync develops. “We’re not leaving this to chance,” Voss said. “The co-founders have skin in the game—their startup’s success depends on our AGI progress, and vice versa. That’s stronger alignment than a traditional acquisition, where the team might check out once the deal closes.”

The Talent War Context

Voss’s defense comes amid an intensifying war for AGI talent. According to a 2025 report from AI industry group OpenAIRE, the number of qualified AGI researchers worldwide is fewer than 500, and top candidates can command salaries of $1–2 million annually, plus equity packages. Amazon AGI Labs, launched in 2023 to compete with Google DeepMind and OpenAI, has struggled to recruit at the same pace as its rivals—partly because it lacks the “startup cachet” that draws researchers to smaller firms like NeuralSync.

The reverse acquihire, Voss said, is a way to “level the playing field.” “NeuralSync’s team didn’t want to leave a startup environment entirely—but they also wanted access to Amazon’s resources: our supercomputing infrastructure, our data sets, and our ability to scale AI safely,” he said. “This structure gives them the best of both worlds.”

Early signs suggest the move is working. Since joining AGI Labs in August, NeuralSync’s research team has already made progress on Amazon’s “ReasonAGI” project, a model designed to help businesses automate complex decision-making (e.g., supply chain logistics, healthcare diagnostics). Meanwhile, NeuralSync’s co-founders have announced their first consumer product: a personal AI assistant called “SyncMind” that uses Amazon’s AGI tech to help users with tasks like planning travel and learning new skills, set to launch in early 2026.

What This Means for Future AI Deals

Voss believes the reverse acquihire could become a more common model in AI, especially as startups and big tech firms seek to balance talent retention with innovation. “The old playbook—buy a startup, shut it down, absorb the team—doesn’t work for AGI,” he said. “These are not just engineers; they’re visionaries. You have to meet them halfway if you want to keep their creativity alive.”

Critics remain cautious, but some VCs are already rethinking their stance. Bennett, of Foundry Group, acknowledged that the structure “could make sense for AI startups where the team’s expertise is more valuable than the IP itself.”

For Amazon AGI Labs, the success of the reverse acquihire will ultimately depend on results: whether the NeuralSync team accelerates Amazon’s AGI progress, and whether the partnership with NeuralSync leads to successful consumer products. But for now, Voss is confident the risk was worth taking.

“AGI is too important to let outdated acquisition models slow us down,” he said. “We’re not just building technology—we’re building partnerships. And this one has the potential to change how AGI is developed and delivered to people.”