Anthropic’s Daniela Amodei on trust as the AI accelerant

Known for its safety-first approach to large language models (LLMs), Anthropic’s influence on the broader uptake of AI has grown exponentially – a trajectory further highlighted by Anthropic’s plans to sell stock to the public and recent valuation of US$956 billion after its latest round of fundraising. 

For enterprise leaders looking to deploy AI reliably at scale, Anthropic’s Claude models have become standard-setting tools.

But, as organisations rush to capitalise on these advancements, the question of balancing the pressure for rapid deployment with the necessity of risk mitigation arises. 

According to Daniela Amodei, Co-Founder and President of Anthropic, the answer doesn’t lie in choosing safety over speed. Instead, true velocity requires building on a foundation of reliability. 

When addressing the critical question of how to balance speed with safety, and what potential pitfalls organisations should keep top of mind, Daniela reframes safety not as a roadblock, but as a catalyst.

“Anthropic has always aimed to be very open about what the drawbacks and challenges are of AI,” she said onstage during the first keynote of Snowflake Summit 26.

 “Part of why we’ve chosen to primarily build for businesses and partner with Snowflake is the concept that trust is an accelerant. 

“Trust is something that helps you go faster.”

“I’ve never had a customer meeting where the CEO said to me ‘I would love if Claude could hallucinate more’,” Daniela added. 

“They’ve also never said it would be great if Claude was less predictable and better at producing bad outputs.

“So, I think the concept that sometimes it’s important to do the safety work and build trust with your customers so you can go faster - that should be an accelerant. I think that’s so much of the foundation that both Snowflake and Anthropic are built on.”

A conversation on stage at Snowflake Summit

This philosophy took centre stage at the Snowflake Summit, which is taking place 1-4 June,  where Snowflake CEO Sridhar Ramaswamy invited Daniela out for an executive Q&A. 

Amid a backdrop of intensifying enterprise partnerships, the conversation quickly turned to the sheer velocity of the market and how much has shifted in the span of just 12 months.

When asked by Sridhar what has changed in one year and what excites her about what organisations can achieve today, Daniela reflected on the condensed timeline of the AI ecosystem.

“In AI, time is this crazy construct – a year ago feels like 10 years ago, right? Five years ago, nobody was using generative AI or LLMs in their daily workflows in their businesses and now every major enterprise says it is a foundational part of their workforce strategy,” she said. 

“Every industry around the world is using AI. I really think a big part of the reason why that’s true is just how much better the models have got in the course of the past year. I think the part that is most mind boggling even to us in Anthropic is what are things going to look like a year from now?”

Though Daniela couldn’t address certain topics at the Summit – jokingly telling Sridhar, “I don’t even know what you’re talking about” regarding the announcement of the firm’s initial public offering – she offered clear guidance on how enterprise buyers can navigate a landscape defined by rapid, compounding capability gains.

Planning for the reality of scaling laws

For IT leaders and C-suite executives, a primary friction point is future-proofing. 

How do you construct a three-to-five-year workforce or technology strategy when the underlying models transform every few quarters?

Sridhar pressed into this issue, asking how people should think about planning and charting their future initiatives in an environment experiencing such aggressive progress. 

Daniela pointed directly to the predictable nature of model development as the key to long-term planning.

“So I love that you are asking about this concept of scaling laws, right? This idea that there’s this kind of predictable way that if you give the models more compute, more data, they get smarter, they get better,” she explained. 

“I think even for those of us at Anthropic where we live and breathe these models and we spend all our time charting them on these graphs and thinking where are they going to be a few years from now, it’s still really hard for us to internalise just the level of capability that these models are developing every year, every six months, every three months.

“And so the advice that we tend to give to customers is to dream big, think about what is the best version of my product or my company or what I want to build for my end customer, because these models truly are developing at such a graphic clip and of course maybe that might change one day, but everything that we have seen indicates that the rate of progress is going to continue.

“And so when you’re thinking about what do I really want to do with AI, don’t just think about today. Also think about what is the biggest version of what I want to build and how do I start building towards that?”

The takeaway for enterprise leaders

For the enterprise tech sector, the alliance between data platforms like Snowflake and AI pioneers like Anthropic signals a mature shift in the market. 

The mandate for businesses is no longer just to build quickly, but to build with a long-term horizon in mind. 

By treating trust as the baseline infrastructure, enterprises can confidently build toward the “biggest version” of their vision, knowing that as the scaling laws march forward, their foundations will remain secure.