Wiz Report Finds AI Services in Use Across More Than 70% of Cloud Organizations
Summary
Wiz’s analysis of hundreds of thousands of cloud accounts found widespread use of AI services and tools, including managed AI services in over 70% of organizations. The report recommends improving visibility and security ownership as adoption grows.
Key points
- The analysis covers aggregate data from hundreds of thousands of cloud accounts across major public cloud providers.
- More than 70% of organizations used managed AI services; 69% had self-hosted AI SDKs and tools, and 42% self-hosted AI models.
- Azure AI Services, including Azure OpenAI, was used by 39% of organizations; deployed Azure OpenAI instances grew 228% over four months in 2023.
- The report found 32% of organizations had fewer than 10 AI service instances, while 10% had 50 or more.
- Wiz recommends tracking AI service usage to reduce visibility gaps, including so-called Shadow AI, and fostering shared security ownership across teams.
Article Details
- Publisher
- Wiz Research
- Report Period
- Includes a four-month measurement period in 2023; report published in 2024.
- Scope
- Aggregate analysis of AI service and tool usage across cloud environments, spanning major public cloud providers.
- Sample Size
- Hundreds of thousands of cloud accounts across a large sample of organizations; exact count not provided.
- Key Statistics
- Over 70% of organizations use managed AI services.
- Self-hosted AI SDKs and tools were found in 69% of cloud environments; 42% of organizations self-hosted AI models.
- Azure AI Services, including Azure OpenAI, was used by 39% of organizations.
- Azure OpenAI deployments rose by 228% over a four-month period in 2023.
- 32% of organizations had fewer than 10 AI service instances, while 10% had 50 or more.
- Recommendations
- Improve visibility into organizational use of AI services and tools to identify blind spots such as Shadow AI.
- For multi-tenant services using generative AI models, follow guidance to protect customer data.
- Foster security ownership and collaboration among security teams, developers, cloud engineers, data scientists, and AI practitioners.