Software companies and SaaS vendors are taking a hit due to the rise of AI agents from OpenAI and Anthropic. Is the SaaS model at risk of disappearing, and what are the consequences?
On February 4, a true ‘SaaSpocalypse’ took place on Wall Street. In just a few days, no less than 400 billion dollars in market value went up in smoke. Major software companies like Microsoft, Oracle, ServiceNow, Salesforce, and Adobe faced a difficult time as investors dumped their shares en masse.
The stock market crash is a result of the rise of AI agents. These have been heralded for some time, but now the agents’ capabilities are getting eerily close to what traditional software vendors offer. Companies are reconsidering whether they still need so many subscriptions and whether AI can take over the work. Thus, the SaaS model suddenly faces a high-stakes re-examination.
Main culprit Anthropic
For what happened on February 4, software vendors can look angrily toward Anthropic. Earlier this year, Anthropic introduced Claude Cowork: an AI assistant focused on various office tasks. Claude Cowork is built on the same foundation as Claude Code, which already made a significant impression with its coding skills.
While Claude Code may only pose a threat to (inexperienced) software programmers, Anthropic is expanding its playing field with Cowork. It brings Claude into the territory of less technically inclined office workers: target groups that SaaS vendors often target with specialized software packages. Claude Cowork bundles all that knowledge into a single platform.
Alongside the rise of Claude Code is OpenClaw, the open-source platform causing a stir. With OpenClaw, you can link an AI agent of your choice to WhatsApp or Signal. OpenClaw is spreading rapidly and theoretically offers everyone the opportunity to build their own AI assistant without needing a subscription. In this article, we delve deeper into the rise of OpenClaw.
SaaS bubble
The SaaS model emerged in the early 2000s. The promise was very simple: for a monthly or annual subscription fee, you get ready-to-use software available via a web browser. Matters such as infrastructure, maintenance, updates, and (to a certain extent) security are all handled by the vendor.
The SaaS market has exploded over the past 25 years and is currently estimated at more than 250 billion dollars. Large organizations utilize on average more than a hundred applications for a diverse range of application areas, from communication to HR and IT security. The period of long growth now seems to be coming to an end.
Current AI tools gradually seem capable of solving the same problems at a lower cost. A paid ChatGPT or Claude account can replace multiple SaaS subscriptions. That is not to say that SaaS vendors are not jumping on the AI bandwagon: Salesforce is claiming the term ‘AI agent’. But where cloud platforms were once the hip alternative to classic software, AI companies like OpenAI and Anthropic now represent the new generation.
To remain relevant in an AI future, software companies will have to reinvent their business model. SaaS software pricing is slowly evolving toward price-per-use instead of per-user. Software companies will also need to leverage the power of their ecosystem as a distinguishing factor to stand out from ‘generic’ AI agents. Salesforce, for example, does this by embedding its Agentforce platform into customer data.
Wanted: seniors
AI has been putting pressure on the job market for some time. Although industry experts claim that AI does not pose a threat to jobs, we are currently seeing quite the opposite. Large technology companies, such as AWS and Workday recently, are announcing major rounds of layoffs or pausing recruitment in the name of AI.
You especially don’t want to be a newly graduated software developer these days. AI coding tools are at a point where they can take over much of the ‘manual labor’ that used to be entrusted to juniors. The role of the developer is changing from programmer to controller, and for this, companies are primarily looking for seniors. Statistics from the US job market show that the employment rate among newly graduated IT professionals is effectively declining.
This trend threatens to cause problems in the long run. If juniors no longer get the chance to gain experience, there will eventually be no more seniors. The knowledge and experience of senior developers thus risks disappearing when they retire. The situation is comparable to mainframes: those large computer boxes are still commonly present today, but no programmer is still proficient in the old Cobol programming language.
Where is it heading?
Although stock market fluctuations are rarely a reflection of rational behavior, the events of February 4 serve as a wake-up call for the software sector. The downward trend has been going on for some time. Investors fear that broader AI adoption will cause a bloodbath among software companies and extend that impact to other sectors.
AI agents are thus shaking not only software but the entire economic system to its foundations. The drive for efficiency gains can hurt in the short term if an entire sector disappears and there is no longer room for junior profiles to develop into seniors. Remove many blocks at the bottom and the Jenga tower will fall sooner or later.