Goldman Sachs warns AI is disrupting software

Goldman Sachs is advising investors that "agentic AI" tools, which can autonomously plan and execute complex tasks, pose a risk to traditional software business models. These AI agents combine large language models (LLMs) with external tools to handle unstructured data and multi-step workflows. The bank's analysts suggest that if AI agents become the primary interface for executing work, traditional software platforms could become passive data stores, eroding their pricing power. This concern has triggered a re-evaluation of software stocks, with investors expecting slower revenue growth. Goldman Sachs has even created an "AI impact framework" to help investors assess the risks and opportunities that AI presents to software companies. This framework considers factors like "orchestration risk," which is the possibility that AI agent layers can bypass traditional platforms. Despite the risks, Goldman Sachs doesn't believe "software is dead". They anticipate AI will stabilize key metrics and pave the way for multi-year growth, especially for companies that innovate and develop differentiated AI-native software. To profit from this shift, Goldman Sachs has launched a pair trade strategy, taking long positions in companies benefiting from AI infrastructure and shorting those vulnerable to AI automation. Short positions include companies whose operations are easily replaceable by AI, reflecting concerns about their future performance.