A new report from Forrester, a research firm that companies pay to help them decide where to spend their technology budgets, has put numbers behind something a lot of executives have felt but couldn't quite prove: AI is not disrupting the software industry evenly. Some corners are about to shrink fast. Others are about to grow just as fast.
Forrester looked at more than 200 different technology and service markets and sorted them by how exposed each one is to AI taking over the work. The ranking runs on nine factors, but the simplest one to understand is this: if a company built its whole business around solving a problem, and AI can now solve that same problem more cheaply, that company is in the danger zone.
Three areas came out as clear winners: cloud computing, cybersecurity, and data management. These are the behind-the-scenes services every AI tool depends on to run, store information, and stay secure, so demand for them keeps climbing no matter which AI tools win or lose on top of them.
The losers are more visible. Custom software development, IT implementation work, and business consulting are all in the direct path of AI that can now write code, run technical setup work, and draft strategy documents on its own. Contracts to set up systems from vendors like Oracle, Salesforce, SAP, and Workday are already showing signs of shrinking work, and outsourced technology and consulting firms are feeling the same pressure on the lower end of their business.
This is already showing up in how investors treat these companies. Software stocks tied to subscription pricing have taken sharp hits over the past year on fears that AI agents will bypass the software entirely and go straight to doing the work, with hundreds of billions in stock value wiped out in a matter of days during some of the worst selloffs.
There is a cost angle too, and it cuts against the idea that AI automatically saves money. A separate study found that the cost of running AI coding tools will pass the salary of an average human developer by 2028, because the price of the computing power behind these tools keeps rising as usage grows. A separate survey of IT teams found that most companies still cannot see clearly what they are spending on AI tools, and a majority reported that wasted AI spending had gone up over the past year.
Put together, this means two things for anyone who manages a technology budget. First, do not assume every software renewal this year deserves the same price next year, because features that once justified the cost may already be commodity work an AI tool can do for less. Second, do not assume switching to AI automatically cuts costs, because without real visibility into usage, AI spending can grow just as wasteful as the software subscriptions it was supposed to replace.