Edge artificial intelligence (AI) adoption is rapidly expanding across industries, according to research presented by ZEDEDA and conducted by Censuswide. In, fact, 97% of CIOs have edge AI either already deployed or on their roadmap, with only 3% reporting no current plans to implement these technologies.
“The findings confirm what we're seeing across industries—edge AI is no longer just a future consideration but an essential component of digital transformation strategies today,” says Said Ouissal, CEO and founder of ZEDEDA. “As a natural evolution of edge computing, edge AI enhances the data processing capabilities already made possible at the edge, enabling organizations to significantly improve customer experience, operational efficiency, and security in ways that cloud-only approaches cannot match.”
Key takeaways:
· 30% of organizations have fully deployed AI at the edge, 22% are actively in production with limited deployment, and 34% are testing with plans to deploy within the next 24 months.
· This widespread adoption spans industries, with retail (50%) and manufacturing (40%) leading in full deployments.
· Customer experience applications currently dominate edge AI implementations, with 80% of CIOs deploying edge AI for use cases that enhance the customer experience, like retail store operations, display personalization, and quality control.
· Risk management applications follow closely at 77%, including predictive maintenance, safety compliance, anomaly detection, and physical security.
· Planned deployments for 2025-2026 show a shift in priorities, with cost reduction (74%) and risk management (73%) leading future implementations. This indicates that organizations increasingly focus on operational efficiencies and risk mitigation after initial customer-facing deployments.
· 93% of retail CIOs implementing edge AI for customer experience, compared to 80% across all industries. Manufacturing strongly focuses on process acceleration for future deployments (82% vs. 68% overall), highlighting industry-specific optimization needs.
· The survey reveals that multimodal AI is the most commonly deployed AI model at the edge (60%) and in the cloud (59%). This indicates that organizations are seeking comprehensive AI solutions that can process and analyze data across multiple formats simultaneously.
· While large language models (LLMs) are equally popular in the cloud (59%), they see somewhat less adoption at the edge (47%), which could reflect computational requirements or use case needs.
· In retail, CIOs reported lower interest in edge-deployed LLMs (32%) but higher adoption of multimodal AI (68%).
· Both driver and challenge security considerations play a dual role in edge AI adoption. Improving security and data privacy is the primary motivation for edge AI investments (53%), followed by improving customer experience (42%) and optimizing operational efficiency (39%). However, security risks and data protection concerns also represent the biggest implementation challenge (42%), followed by high operational and maintenance costs (40%).
· Other significant challenges include finding the right technology vendors and partners (37%) and a shortage of talent with edge AI expertise (37%).
· Most organizations (54%) report that edge AI complements their cloud AI strategy for a hybrid approach. Nearly half (48%) are exploring edge AI specifically to reduce cloud computing costs, while 44% consider edge AI critical for real-time processing and low-latency requirements.
· 90% of CIOs reporting increases for 2025. Three in 10 (30%) organizations are significantly increasing edge AI budgets by 25% or more, while 60% report moderate increases of up to 25%.
· Larger businesses (500-plus employees) are more aggressive in their investments, with 39% reporting significant budget increases compared to 23% of mid-sized organizations (250-500 employees).
