AI Spending Set to Surge by 2027 Despite ROI Struggles

Artificial intelligence is poised to consume a fifth of all business technology spending by 2027, as The Register first reported. However, this increase in capital is not translating into broad operational efficiency.

Forrester's State Of AI Survey 2026 reveals a significant gap between investment and returns. Only 22% of AI decision-makers report a positive impact on EBITDA over the last year. Furthermore, Forrester expects more than half of CIOs to miss their AI-related ROI targets in the coming year, noting that success is likely to be limited to isolated use cases rather than comprehensive programs.

According to the report, the lack of measurable, firmwide results stems from a failure to update organizational structures, processes, and workflows to match the technology adoption. Sharyn Leaver, Forrester's chief research officer, noted that this creates pressure for leaders to demonstrate value while managing risks and infrastructure costs.

To mitigate these costs, Forrester predicts a shift toward open-weight models like Qwen, Llama, and Gemma. These models are deemed sufficient for high-volume enterprise tasks such as internal knowledge search, customer support, and document processing. Forrester estimates open-weight models will account for 50% of AI tokens but only 5% of enterprise spending. In contrast, high-cost frontier models—including Google's Gemini 4 Argon, OpenAI's GPT-6 Astra and GPT-6.1 Sol, and Anthropic's Claude Sonnet 5.5 and Claude Opus 5.5—will be reserved for critical tasks where errors are costly.