On Friday, Google introduced a partnership with Anthropic, a startup based in 2021. Anthropic makes Claude, which like ChatGPT is a synthetic intelligence chatbot. 

With the $300 million deal, Google will take a roughly 10% stake within the startup, in keeping with the Monetary Occasions. Anthropic for its half will get each a monetary enhance and cloud computing assets it wants. The deal may give Anthropic a valuation of roughly $5 billion, the New York Occasions reported.

In fact, ChatGPT maker OpenAI has Microsoft cash behind it—plenty of it. In a deal revealed final month, Microsoft will make investments $10 billion within the startup. It additionally invested $1 billion in 2019, then quietly added one other $2 billion in 2021. The software program large plans to include OpenAI know-how into all kinds of merchandise, together with the Bing search engine, probably disrupting Google’s search dominance.

The 2 startups share some DNA. Anthropic was fashioned principally from a bunch that broke away from OpenAI after rising disillusioned with strategic and cultural shifts. One former OpenAI worker not too long ago advised Fortune that whereas OpenAI founders Sam Altman and Greg Brockman paid lip service to A.I. security, “that usually appeared like only a fig leaf for enterprise considerations, whereas precise, reputable A.I. security considerations have been brushed apart.” 

Anthropic says it’s centered on A.I. security—it calls itself “an A.I. security and analysis firm”—and factors to its analysis on “Constitutional A.I.”

With the latter, it writes on its web site, “We present that language fashions can be taught to observe a set of straightforward, pure language rules by way of self-improvement, and we use this new methodology to coach a extra innocent assistant.” 

Tesla CEO Elon Musk has warned in regards to the risks of superior A.I. In December he referred to as ChatGPT “scary good,” including, “We’re not removed from dangerously sturdy AI.” He additionally tweeted that his confidence in OpenAI’s security was “not excessive,” noting that, with assist from him, it “began as open-source & non-profit. Neither are nonetheless true.”

Fortune reached out to OpenAI for remark however didn’t obtain a right away reply.

Anthropic launched a restricted take a look at of Claude in January and plans to share it with a wider viewers. 

“We’re partnering with Google Cloud to help the following section of Anthropic, the place we’re going to deploy our AI programs to a bigger set of individuals,” Anthropic CEO Dario Amodei mentioned in a press release. “This partnership offers us the cloud infrastructure efficiency and scale we want.”

He added, “We’re keen to make use of the Google Cloud infrastructure to construct dependable, interpretable, and steerable A.I. programs.”

Learn to navigate and strengthen belief in your small business with The Belief Issue, a weekly publication inspecting what leaders must succeed. Enroll right here.