Check Point's 2026 report: navigating AI challenges in cloud security

In its latest release, Check Point Software Technologies examines the challenges faced by businesses amid the rapid adoption of AI technologies in its 2026 Cloud Security Report: Enter the AI Era. The report describes a gap between the adoption of AI and existing security measures, which can create vulnerabilities in cloud environments.

The report outlines a shift from 2025’s cloud “blind spots” to governance, control, and real-time enforcement issues in 2026. As AI changes user behaviour and application communication, it introduces new security risks. It notes that 77% of organisations have adjusted their cloud security strategies to include AI this year, while 26% feel able to enforce these strategies effectively. This 51-point difference highlights the gap between planned approaches and implementation capability.

As organisations work to use AI capabilities, cybercriminals are also using these tools to support phishing schemes, generate more advanced malware, and increase the speed of attacks. The report also states that 78% of companies reported experiencing confirmed or suspected AI-related security breaches in the last year.

According to the findings, cloud-native environments face several challenges:

• Infrastructure Misalignment: 52% of AI workloads span hybrid environments, while 64% of respondents indicate a need for architectural redesign.
• Perimeter Gaps: 76% consider data centre security important for AI, while 35% of systems currently meet these requirements.
• Performance Issues: 24% can inspect AI traffic without performance impact, while 71% report an increase in WAF false positives.
• Operational Complexity: 88% report increased security complexity due to AI adoption, while 67% note fragmented policy enforcement.
• Limited Visibility: More than half of organisations have experienced AI-related security incidents, while 24% report uncertainty due to visibility limitations. This suggests that a large proportion may still face exposure.
• Identity Risks: Non-human identities, including AI agents and APIs, are identified as a key area of concern.
• Access Model Inconsistencies: 24% of organisations report not having AI-specific access controls, while 16% apply consistent security measures.

In response to these challenges, Check Point outlines a prevention-focused architecture across cloud, data centre, SaaS, and endpoint environments. Its Hybrid Mesh Network Security approach includes:

• Unified Management: A hybrid mesh architecture designed to support consistent policies and protections.
• Prevention-First Security: AI-driven approaches intended to block threats such as ransomware, with a reported 99.8% effectiveness in recent evaluations.
• Secure Connectivity: Identity-based, real-time protection for users, devices, and applications.
• AI Defense Plane: Centralised control of AI deployment and operation, providing runtime protection.
• Agentic Network Security Orchestration: Aims to reduce operational complexity, align with business intent, and support Zero Trust principles.