---
schema_version: "1"
id: 5a5bc78ebcfa1839f1b74bf39ac06a8244df0b134518bcdfe418744fde331192
title: Conduct a Risk Assessment for AI Projects
collection: admin
version: current
locale: en-us
portal_url: https://docs.salesforce.com/admin/current/en-us/ai/264-0-0/generative_ai_plan_trust.md
canonical_url: https://help.salesforce.com/s/articleView?id=ai.generative_ai_plan_trust.htm&release=264.0.0&type=5
source_site: help.salesforce.com
content_type: documentation
content_hash: e71c489f463b6b78a9fcc794ad00073d66f1361ef971940600d07b9c349e2480
description: Identify the risks of using data with AI solutions and establish guardrails to keep customer data safe.
products:
  - Agentforce
  - Cross Cloud
release: "264"
---
# Conduct a Risk Assessment for AI Projects

Identify the risks of using data with AI solutions and establish guardrails to keep customer data safe.

Using customer data in your prompts results in better responses from the model. However, including business data in your prompts means that you’re opening up your company to risks. Data leaks, regulatory issues, and reputational harm can happen if your data isn't handled with care.

After identifying the data that your AI project requires, you can establish risk mitigation strategies that are designed to keep that customer and business data secure.

1. Work with your project stakeholders to identify the likelihood of negative effects of each of these risk types with the project data. You can use the [Plan Your Trust Strategy](https://trailhead.salesforce.com/content/learn/modules/ai-data-project-planning/plan-your-trust-strategy) unit in Trailhead to build a risk profile and mitigation strategy to share with key stakeholders.
2. Identify guardrails that can help your project mitigate legal and ethical risks. Many of these guardrails are covered by the [Einstein Trust Layer](https://help.salesforce.com/s/articleView?id=ai.generative_ai_trust_layer.htm&release=264.0.0&type=5 "Trusting generative AI with your company secrets and your customer data while balancing the benefits of generative AI is still a challenge today. The Einstein Trust Layer is a secure AI architecture, built into the Salesforce platform. It’s a set of agreements, security technology, and data and privacy controls used to keep your company safe while you explore generative AI solutions."). AI guardrails include:
    1. Security guardrails: Secure data retrieval means the prompt uses only data that the end user is allowed to access. Data masking replaces sensitive data with placeholder data so the model never sees the masked data. Zero data retention policies mean that data isn’t stored in the model after the response is generated.
    2. Technical guardrails: Prompt Defense protects the data from prompt injection attacks or jailbreaking.
    3. Ethical guardrails: Toxicity and bias detection identifies harmful language in prompts and responses.
3. Build a list of potential risk areas and the specific guardrails that the system uses to mitigate risk.
4. Prioritize the risk mitigation strategies so that you tackle the most critical risks first.

#### See Also

- [_Trailhead_: AI + Data: Project Planning](https://trailhead.salesforce.com/content/learn/modules/ai-data-project-planning)
- [_Trailhead_: AI Strategy](https://trailhead.salesforce.com/content/learn/modules/ai-strategy)