Anthropic on Tuesday announced a new suite of enterprise offerings, the latest in a series of new capabilities boasted by AI labs like Google, Anthropic, and OpenAI that have shaken investor confidence in a variety of software solutions that may see their core product replaced by an AI facsimile.
But some Wall Street analysts aren’t so sure the process will be quite that simple.
Dan Ives, high profile tech bull at Wedbush, wrote in a note Tuesday after Anthropic’s announcement that, “While these use cases are impressive, the reality is that these new AI tools will not rip and replace existing software ecosystems and data environments with these AI tools only as useful as the data it can reach.”
In Ives’ view, there are three things investors who view these demos as a sign of AI outcompeting legacy solutions might be missing.
One — “The market is conflating foundation model capability with full enterprise software replacement and the fear that generative AI will ‘rewrite legacy systems overnight’ ignores enterprise reality.”
Two — “AI agents and autonomous workflows dramatically increase the attack surface — more APIs, more machine identities, more lateral movement risk, and more cloud-native workloads. AI doesn’t reduce the need for endpoint, identity, cloud, and SOC automation, it multiplies it.”
Three — “Anthropic/OpenAI do not have 20-year enterprise distribution networks, CIO relationships, or embedded vertical workflows. CRM, NOW, and MSFT sit at the application layer where business logic lives. The model layer will commoditize faster than the workflow layer.”
The key point that I think Ives raises is that model advances — or other signs of what we might call technical brilliance — are not the same as an in-market product that gets customers to change their spending habits.
Obviously, as we saw with the market’s reaction to a Substack post on Monday that was explicitly a thought experiment, there are lots of investor nerves about which advances could draw meaningful dollars from enterprises or consumers.
But as ever, the distance between a demo and a product can often be vast. Keeping that framework in mind might serve investors well in the current moment.
