> ## Documentation Index > Fetch the complete documentation index at: https://docs.planeconnection.com/llms.txt > Use this file to discover all available pages before exploring further. # How to Use the Safety AI Assistant > Ask safety questions, analyze patterns across reports and investigations, get AI-generated summaries, and receive actionable recommendations. By following this guide, you will use the Safety AI Assistant (ALI) to analyze safety data patterns, draft investigation content, and receive actionable recommendations grounded in your workspace's records. **Who should read this:** Safety managers, investigators, and admins who want to accelerate safety analysis and decision-making. Any workspace member with access to the Safety module can use ALI. **Prerequisites:** Access to the Safety module. ALI is available from the Safety sidebar and from within individual reports and investigations. ## Open the AI Assistant There are two ways to access ALI: * **From the Safety sidebar.** Click **ALI** in the Safety navigation to open the full chat interface. * **From a record.** Click **AI Assist** in the header of a report, investigation, or CPA detail page. The assistant opens with context about that specific record. When opened from a specific record, ALI automatically has context about the record number, status, findings, and related data. This enables more targeted analysis without re-entering details. ## Ask Safety Questions Type your question in the chat input. ALI can help with a wide range of safety management tasks: | Task | Example Prompt | | -------------------- | ------------------------------------------------------------------------ | | Pattern analysis | "What are the most common hazard types reported in the last 90 days?" | | Trend identification | "Are unstable approach reports increasing this quarter?" | | Regulatory guidance | "What does 14 CFR 5.71 require for safety performance monitoring?" | | Drafting content | "Draft a summary for investigation INV-2026-00042." | | Risk analysis | "What contributing factors are common in our ground handling incidents?" | | Comparison | "How does our report submission rate compare to the target SPI?" | Be specific in your prompts. "Analyze maintenance reports from Q4 2025" produces more useful results than "Tell me about maintenance." Include time ranges, report types, or categories to narrow the analysis. ## Review AI-Generated Citations ALI grounds its responses in your actual safety data. Responses include citations that reference the source records: | Citation Type | Example | | ------------- | --------------- | | Report | RPT-2026-00123 | | Investigation | INV-2026-00042 | | CPA | CPA-2026-00015 | | Document | SMS Manual v3.2 | | Regulation | 14 CFR 5.55 | Click on any citation to navigate directly to the source record. This lets you verify the AI's analysis against the original data. **AI suggestions are advisory only.** Per 14 CFR Part 5, all safety assessments, investigation findings, and risk evaluations require independent human evaluation by qualified personnel. AI recommendations should be used as one factor among many in decision-making, not as the sole basis for any safety, operational, or maintenance action. See [Legal Notices](/en/reference/legal-notices#ai-and-automated-features-disclaimer). ## Use Action Suggestions When ALI identifies potential actions, it presents them as clickable action cards with direct links to the relevant module: * **Create a report** -- opens the new report form with pre-filled context. * **Open an investigation** -- navigates to the investigation creation workflow. * **Create a CPA** -- opens the CPA form linked to the relevant finding. * **Start a risk assessment** -- opens the risk form with identified hazard details. Action suggestions are recommendations, not automated actions. You decide whether to follow through. ## Analyze Patterns Across Reports Use ALI to identify recurring themes and systemic issues that may not be visible in individual reports: 1. **Ask about patterns.** For example: "Are there recurring hazards in our flight operations reports from the last 6 months?" 2. **Review the analysis.** ALI examines report types, categories, locations, aircraft, and contributing factors to identify clusters. 3. **Follow up with specifics.** Ask narrowing questions such as "Which airport codes appear most frequently in those reports?" or "What time of day are these events occurring?" 4. **Take action.** If patterns suggest a systemic issue, use the action suggestions to create a risk assessment or investigation. ## Draft Investigation Content When working on an investigation, open ALI from the investigation detail page to: * **Draft summary language** -- ALI generates a narrative summary based on the investigation's findings and root causes. * **Identify contributing factors** -- suggests additional factors to consider based on similar investigations. * **Suggest lines of inquiry** -- recommends additional questions or evidence to pursue. Use ALI to draft initial content, then review and refine it with your professional judgment. AI-generated text should be a starting point, not the final product. ## Natural Language Reporting ALI powers the natural language report submission mode. When you describe a safety event in plain text, the AI extracts structured fields -- report type, severity, category, location, date/time, and aircraft -- for your review. See [Submit a Safety Report](/en/how-to/sms/submit-safety-report) for the full natural language workflow. ## Best Practices * **Verify all outputs.** Cross-reference AI analysis against source records before including in formal documents. * **Iterate on prompts.** If the first response does not fully address your question, refine your prompt with additional context or constraints. * **Use contextual mode.** Open ALI from within a specific record to get more targeted, relevant analysis. * **Document AI-assisted analysis.** When AI insights inform a safety decision, note that AI analysis was used and describe the independent verification you performed. ## Related Natural language report submission powered by ALI. Using AI Assist during investigations. SPI monitoring that triggers AI-assisted pattern analysis. Risk assessments informed by AI pattern identification.