人工智能方案目錄

Palo Alto Networks AI Access Security provides a unified platform to discover, monitor, control, and secure generative AI applications. Rather than completely blocking AI, it enables a Zero Trust security posture for AI adoption across the network, using the following key architecture pillars:

1. Comprehensive Visibility & Risk Profiling: Automatically discovers shadow and sanctioned GenAI tools in use across departments. It evaluates hundreds of GenAI apps against custom risk-scoring models to empower department cybersecurity teams to make data-driven decisions. 2. Granular Access Control: Rather than all-or-nothing bans, IT managers can configure policy rules to restrict functionality (such as allowing text prompts but blocking file uploads or custom browser plug-ins) based on the risk tier of the application. 3. Precision AI Data Protection (Enterprise DLP Integration): Built entirely upon Palo Alto Networks' class-leading Enterprise Data Loss Prevention (E-DLP) engine, this solution uses over 300 data classifiers, context-aware machine learning (ML), and LLM-powered classification. It continuously monitors prompt traffic inline to identify and block the transmission of personal data (PII), proprietary code, financial figures, and sensitive internal records. 4. Real-Time Threat Prevention & Safety: Scans the specific AI payloads (text responses, files, and code snippets) from sanctioned and allowed apps. This blocks system-level exploitation attempts, malicious URLs, and malware before they enter the internal network.

Use Case 1: Securing AI-Powered Citizen Enquiry Portals / Securing AI-Powered Workforce The AI Access Security Solution: AI Access Security inspects the responses from both sanctioned and tolerated GenAI applications. It screens those incoming files, URLs, and code snippets returned by the LLM in real-time, leveraging Precision AI security services to stop malware and malicious links before they hit civil servant browsers.

Use Case 2: Safeguarding Citizen PII During Licensing & Permit Triage The AI Access Security Solution: The solution utilizes high-performance Enterprise DLP fueled by machine learning to scan text prompts and file uploads inline. If a user tries to upload document PDFs containing citizen PII (like HKID patterns or names) to an AI tool, the platform intercepts and blocks the transfer instantly. It provides fine-grained, context-aware policy controls, allowing officers to use GenAI for basic text generation but restricting file uploads or browser extensions.

Use Case 3: Controlling "Shadow AI" Across Complex Policy Research and Design The AI Access Security Solution: AI Access Security discovers, categorizes, and monitors thousands of GenAI applications across the whole estate, assigning bespoke risk scores to each tool. It lets IT administrators transition from blanket-blocking AI to orchestrating highly granular access rules via a unified dashboard