Multimodal AI is the most commonly deployed AI model at the edge (60%).

Compared to publicly available tools, 63% agree that domain-specific AI significantly or extremely enhances security operations.

Large Language Models (LLMs) are widely deployed in the cloud (59%).

34% of organizations are testing with plans to deploy AI at the edge within the next 24 months.

High operational and maintenance costs of Edge AI are a challenge for 40%.

Only 3% of CIOs report no current plans to implement Edge AI.

90% of organizations are increasing edge AI budgets for 2025.

70% of the surveyed government fraud fighters have seen an uptick in AI-powered fraud attacks in the last five years.

A third of respondents indicate that GenAI is either being integrated or is actively transforming their operations.

Security risks and data protection concerns represent the biggest Edge AI implementation challenge (42%).

1 in 3 (33%) of respondents plan to fill skills gaps with AI and automation.

There was a 2,000% rise in malicious sites containing "openai" in their name between April 2024 and April 2025

85% of respondents listed fighting fraud as a top five priority.

Nearly a third face significant resource limitations in fighting FWA.

In retail, CIOs reported lower interest in edge-deployed LLMs (32%).

LLMs see somewhat less adoption at the edge (47%).

31% of SOCs are using GenAI for Querying security data.

Multimodal AI is widely deployed in the cloud (59%).

77% of CIOs with deployed edge AI solutions focus on risk management applications.

The use of network analysis for fraud detection is expected to expand from 32% to 87%.