Here’s Why Anthropic Refuses to Offer 9-Figure Pay Like Meta

While Meta poaches talent from Apple, OpenAI and Google, AI startup Anthropic is refusing to play the game by matching competing offers.

Anthropic CEO Dario Amodei explained his reasoning on an episode of the “Big Technology Podcast,” released earlier this week. Amodei said that he recently sent a Slack message to all Anthropic staff informing them that the company was not willing to “compromise our compensation principles” or its “principles of fairness” when individual employees receive outside offers.

He said that Meta’s efforts to poach staff were a “unifying moment” for the company, citing his decision not to match offers due to potential unfairness for other staff members.

Related: AI Is Dramatically Decreasing Entry-Level Hiring at Big Tech Companies, According to a New Analysis

Amodei also acknowledged on the podcast that fewer Anthropic employees had been captured by Meta’s compensation offers when compared to other companies, though “not for lack of trying.” Some Anthropic staff “wouldn’t even talk” to Meta CEO Mark Zuckerberg, according to Amodei.

Meta is reportedly offering more than $200 million in compensation to one AI researcher on the superintelligence team who worked at Apple. The tech giant did manage to poach Anthropic software engineer Joel Pobar, as of a June 30 memo.

“If Mark Zuckerberg throws a dart at a dartboard and it hits your name, that doesn’t mean that you should be paid 10 times more than the guy next to you who’s just as skilled, just as talented,” Amodei said on the podcast.

^{Anthropic CEO Dario Amodei. Photo by Halil Sagirkaya/Anadolu via Getty Images}

Anthropic’s compensation is tied to a level-based system. Amodei explained on the podcast that when Anthropic staff join the company, they are classified into one of many different levels, which corresponds to their compensation.

“We don’t negotiate that level because we think it’s unfair,” Amodei said. “We want to have a systematic way.”

Related: How Much Does It Cost to Develop and Train AI? Here’s the Current Price, According to Anthropic’s CEO.

Amodei said that Anthropic’s mission of safely creating reliable, cutting-edge AI systems inspired many employees to stay, and asserted that Zuckerberg was “trying to buy something that can’t be bought,” which is alignment with a company’s mission.

Zuckerberg, meanwhile, recently outlined his mission with his superintelligence team, a group working on creating AI that surpasses human intelligence. In a blog post on Meta’s website, published on Wednesday, Zuckerberg said that Meta’s goal was to bring superintelligence to every individual and allow people to reap the creative, economic and personal benefits of the technology.

He contrasted the effort with the intentions of “others in the industry” who want to use AI to automate the workforce first before giving it to individuals. Meta’s mission is to empower individuals with AI, Zuckerberg wrote.

Related: Reddit Sues $61.5 Billion AI Startup Anthropic for Allegedly Using the Site for Training Data

Since its start in 2021, Anthropic has raised close to $20 billion from companies like Google and Amazon. According to a Bloomberg report from earlier this week, the startup is nearing a deal to raise funds at a $170 billion valuation.^{Join top CEOs, founders and operators at the Level Up conference to unlock strategies for scaling your business, boosting revenue and building sustainable success.}