Enterprise AI has moved beyond the experimental phase, but adoption remains uneven across organizations. According to Futurum Group's latest AI Decision Maker Survey, businesses are increasingly adopting multi-model approaches, using different AI models for specific tasks rather than relying on a single provider. This strategy allows enterprises to optimize performance, cost, and compliance.
"We're seeing a clear divide between AI power users and laggards," said Stephen Foskett, analyst at Futurum Group. "Power users are organizations that have integrated AI deeply into workflows, often deploying multiple models simultaneously. Laggards are still stuck with legacy systems and workforce adaptation challenges."
The survey highlights that AI power users are defined not just by the number of models they use, but by their ability to orchestrate agentic AI systems—autonomous agents that can plan and execute tasks. These organizations are moving beyond simple chatbots to AI agents that can handle complex, multi-step processes.
"Multi-model strategies are not the same as multimodal AI," explained Nick Patience, also of Futurum Group. "Multimodal AI processes text, images, and audio together. Multi-model means using different AI engines—like GPT-4, Claude, and Gemini—for different jobs, depending on accuracy, latency, and cost requirements."
The report also notes that legacy systems and workforce adaptation remain the biggest barriers to AI adoption. Companies that fail to modernize their data infrastructure or upskill employees risk falling behind. However, the payoff for those that succeed is significant: power users report 30-40% efficiency gains in key processes like customer service, code generation, and data analysis.
As 2026 progresses, the gap between leaders and laggards is expected to widen, with agentic AI and multi-model strategies becoming the new baseline for competitive enterprises.