AI Agents Are On The Horizon — Here Are The Software Stocks Bernstein Says Will Survive

TLDR

• Bernstein indicates that AI is transforming enterprise software by focusing on “control planes,” which are the infrastructure layers responsible for managing AI capabilities.
• Cloud infrastructure segments, including IaaS and PaaS, are anticipated to experience the most significant benefits from the AI transition.
• While legacy software faces considerable pressure, demand is not vanishing but rather being re-evaluated in terms of pricing.
• Consumption-based pricing models and premium AI bundles are emerging as the primary drivers of future growth.
• Bernstein suggests that the notion of the “death of software” is an exaggeration.

(SeaPRwire) –   Bernstein has presented a five-year forecast detailing how artificial intelligence is expected to reshape the software industry. The firm’s central assertion is that AI is not eliminating software but rather altering the methods by which value is generated and captured.

The report introduces the concept of an “AI control plane,” defined as a fundamental infrastructure layer that oversees and synchronizes AI tools, data streams, and software agents within a business. Bernstein posits that companies controlling this layer will emerge as the long-term victors.

The firm firmly believes that AI is already exerting a tangible influence on enterprise software, rather than being a concern for the distant future.

Cloud Infrastructure Stands to Gain Most

Bernstein anticipates that cloud infrastructure providers, particularly those offering Infrastructure-as-a-Service and Platform-as-a-Service products, will see the most substantial gains. There is a rapidly increasing demand for both GPU and standard computing capacity.

This growth is projected to accelerate as “agentic AI”—software designed to operate autonomously to achieve objectives—becomes more prevalent in large corporations. These systems require substantial backend support to function effectively.

Database utilization is also expected to increase. Bernstein forecasts that more businesses will transition away from on-premise legacy systems towards cloud-based and AI-native database solutions.

The firm views this trend as an expansion of the total addressable market for major technology companies, rather than a contraction.

Not All Software Is Equal

Bernstein distinguishes clearly between software that is susceptible to disruption and software that is resilient. Legacy products reliant on traditional licensing models are facing significant pricing challenges.

The MSCI World Software and Services Index has seen a decline of over 20% this year, reflecting widespread market apprehension regarding AI’s disruptive potential. Bernstein contends that this market reaction has been overly generalized.

The firm argues that AI has not diminished demand for software but has instead led to a repricing of its value. For instance, the IT services sector is shifting from billing based on hours worked to billing based on achieved outcomes.

Companies that can demonstrate that AI features are driving increased usage among their existing customer base will be better positioned to regain investor confidence.

Industry perspectives further support this outlook. Ido Arieli Noga, CEO of Yuki, suggests that AI agents do not replace data platforms but rather extract data from them. He warns that the growing adoption of agents could lead to a surge in infrastructure consumption as AI systems generate queries continuously.

Bernstein identifies a key risk to its own analysis: if the costs of compute power and electricity escalate significantly, the economic viability of expanding AI infrastructure could face substantial limitations.

The firm asserts that the new key financial metrics to monitor are usage rates, the attach rates of AI features, and recurring revenue linked to consumption-based pricing, rather than solely focusing on headline revenue growth.

Bernstein’s report was released on April 19, 2026.