Salesforce Stock Drops 8% as Claude Makes Investors Scared of AI Security Threats


Salesforce stock is down 8% this week, hitting its lowest level since March 2022.

Although multiple factors have appeared to impact the stock’s sudden drop, its timing coincides with an enterprise software mass selloff after Anthropic revealed a preview of its latest AI model, Claude Mythos. This has prompted investors to consider the ramifications of potential AI cybersecurity threats.Ā 

Why Is the Stock Down?

CRM stock has always been turbulent, and this year is no exception. In January, we analyzed why the stock had dropped 25% in just one year; it then slid further down in February, and in March, it finally experienced a rise.Ā 

However, on Wednesday, the stock dropped from $176.86 to $169.34.Ā 

At the time of writing, the stock is currently marginally up to $170.85, having experienced a dip of 8.15% over the last five days.

ā€œSaaSpocalypseā€ claims and debates over whether traditional enterprise software can compete in an increasingly AI-driven market have caused numerous stocks to rise and fall lately, and it is likely one of the factors behind this particular Salesforce drop. However, it isn’t the only one – so what else has spooked investors?Ā 

READ MORE: Huge Agentforce Growth in Salesforce Q4 as Benioff Mocks ā€˜SaaSpocalypse’ Narratives

Enter Claude MythosĀ 

On Wednesday, AI company Anthropic announced a limited release of its latest model, Claude Mythos. It became a topic of conversation almost immediately, as every published benchmark shows that it is the most capable AI model ever built.

Not only that, but it has even been deemed too dangerous to release to the public by its developer due to how proficient it is in finding critical security flaws in computer systems.

When news of Mythos was first leaked at the end of March, enterprise security stocks instantly took a hit, with the likes of CrowdStrike and SentinelOne dropping between 5-11%. Investors were worrying that AI models could undermine demand for traditional security products. Now, at the time of the product’s limited release, the focus (and worry) is back on cybersecurity, but for a slightly different reason.Ā 

ā€œIf Mythos is that strong and that powerful and it’s exposing these vulnerabilities that have been ​around for years, it just shows one, the weakness of the current software that’s out there, and two, that AI ​is still making incredible progress versus legacy software companies,ā€ said Michael O’Rourke, chief market strategist at JonesTrading.

READ MORE: 5 Principles for Salesforce Security in 2026

Salesforce was just one of the multiple enterprise software stocks that dropped after the news, joining the likes of Atlassian, Workday, and Adobe with drops between 3.7% and 6.8%.

Potential AI Cybersecurity Threats and Market Implications

Ultimately, there are two layers to this. Firstly, there is the direct cybersecurity fear. If advanced AI can help locate and weaponize security flaws more easily, investors worry that cyber risk rises for everyone who builds or runs software. Anthropic’s own write-up frames this as a major step-change in capability, which is why access is being limited to a small defensive program rather than opened broadly.

Then there is the business model fear. Investors are likely considering the fact that frontier AI can do more of the hard work that today’s software products charge for – like detecting issues, automating remediation, writing code, or replacing parts of security and enterprise workflows – which means some existing software companies could face faster disruption than expected.Ā 

This has caused a sector-wide selloff effort as investors reevaluate the role that enterprise software has in this existing market. Salesforce, like many of its competitors, offers an AI-driven SaaS solution, but it still pales in comparison to many of the capabilities being demonstrated by these latest AI models. Whether it will be enough to continue succeeding is still unclear.

READ MORE: New Setup With Agentforce: How Salesforce Admins Can Get Started

Final ThoughtsĀ 

As AI gets increasingly more powerful, the possibilities of what could happen if it goes wrong increase tenfold. You could argue that it is surprising AI cybersecurity threats have not been as widely considered yet as they should, but it takes just one error or exposed vulnerability before an org is destroyed from the inside out, and mountains of customer data are leaked.Ā 

For Salesforce, this is not the last time its stock will drop or fluctuate. We are likely to see it rise next week amid the news from its developer conference, TrailblazerDX, hopefully introducing new capabilities to stand out in this competitive market. Ā 

Leave a Reply

Your email address will not be published. Required fields are marked *