August 4, 2025
5 min read
Netskope
Netskope reveals a 50% surge in enterprise genAI platform use, highlighting growing shadow AI risks and the need for enhanced AI governance.
Netskope Threat Labs: Shadow AI Risks Proliferate as GenAI Platforms and AI Agents See Rapid Adoption
Latest research from Netskope shows a 50% increase in generative AI (genAI) platform usage among enterprise end-users in the first half of 2025. Despite efforts to safely enable SaaS genAI applications, shadow AI—unsanctioned AI tools used by employees—continues to grow, posing significant security challenges.The Rise of genAI Platforms
GenAI platforms, which allow organizations to build custom AI applications and autonomous AI agents, are the fastest growing category of shadow AI due to their ease of use and flexibility. Usage of these platforms increased by 50% in the three months ending May 2025. This surge accelerates risks related to enterprise data security, emphasizing the importance of data loss prevention (DLP) and continuous monitoring.- Network traffic linked to genAI platform use rose 73% over the prior quarter.
- In May 2025, 41% of organizations were using at least one genAI platform.
- Microsoft Azure OpenAI leads with 29% adoption, followed by Amazon Bedrock at 22%, and Google Vertex AI at 7.2%. Ray Canzanese, Director of Netskope Threat Labs, stated, "The rapid growth of shadow AI places the onus on organizations to identify who is creating new AI apps and agents, and where they are deployed. Security teams must evolve their AI app controls and DLP policies to include real-time user coaching to safeguard innovation."
- 34% of organizations use LLM interfaces, with Ollama leading at 33% adoption.
- AI marketplaces like Hugging Face are popular, with 67% of organizations downloading resources.
- GitHub Copilot is used in 39% of organizations, and 5.5% run AI agents generated from popular frameworks on-premises.
- Two-thirds (66%) of organizations have users making API calls to api.openai.com; 13% call api.anthropic.com.
- Enterprise users are consolidating around purpose-built tools like Gemini and Copilot, which are better integrated into productivity suites.
- ChatGPT saw its first decline in enterprise popularity since 2023.
- Other apps like Anthropic Claude, Perplexity AI, Grammarly, and Gamma saw adoption growth.
- Grok entered the top 10 most-used genAI apps despite still being among the most-blocked, with blockage rates decreasing.
- Assess the genAI landscape: Identify which tools are in use, by whom, and how.
- Strengthen genAI app controls: Enforce policies allowing only approved genAI apps, implement blocking, and real-time user coaching.
- Inventory local AI infrastructure: Review security frameworks like OWASP Top 10 for Large Language Model Applications and secure data, users, and networks.
- Continuous monitoring: Detect new shadow AI instances and stay updated on AI ethics, regulations, and adversarial threats.
- Address agentic shadow AI risks: Collaborate with early adopters to develop realistic policies limiting shadow AI. For more details, view the Netskope Threat Labs Cloud and Threat Report: Shadow AI and Agentic AI 2025.
- Assessing the current landscape of AI tools being used.
- Implementing strong controls to allow only approved AI applications.
- Conducting real-time user coaching to educate employees on safe AI usage.
- Continuously monitoring for new instances of shadow AI.
- Developing clear policies for the use of AI agents, especially in collaboration with early adopters.
- Ensuring the security of local AI infrastructure by reviewing security frameworks like the OWASP Top 10 for LLMs.
- AI Agents: Capabilities, Risks, and Growing Role - Delve deeper into the landscape of AI agents and their implications.
- AI-Driven Crypto Trading Tools Reshape Market Strategies - Explore how AI is transforming the cryptocurrency trading environment.
- Navigating the Future of Cryptocurrency: What's Changing and Why It Matters - Gain insights into the evolving cryptocurrency market.
On-Premises AI Innovation
Organizations are increasingly deploying genAI locally using on-premises GPU resources and developing tools that interact with SaaS genAI applications. Large Language Model (LLM) interfaces are gaining traction:Evolution of SaaS AI Use
Netskope tracks over 1,550 distinct genAI SaaS applications, up from 317 in February 2025. Organizations now use about 15 genAI apps on average, up from 13, with data uploads increasing from 7.7 GB to 8.2 GB per month.Recommendations for AI Governance and Monitoring
CISOs and security leaders should take proactive steps to ensure safe AI adoption:About Netskope
Netskope is a leader in modern security and networking, providing optimized access and real-time, context-based security for people, devices, and data anywhere. Trusted by thousands, including over 30 Fortune 100 companies, the Netskope One platform offers Zero Trust security and powerful networking to reduce risk and provide visibility and control over cloud, AI, SaaS, web, and private applications. Learn more at netskope.com, on LinkedIn, and Instagram.Source: Originally published at PR Newswire on August 4, 2025.