Salesforce AI Associate Certification Exam Guide
📋 Quick Navigation
⚡ Quick Answer
What is the Salesforce AI Associate Certification Exam Guide?
The Salesforce AI Associate Certification Exam Guide validates expertise in the relevant Salesforce domain. Exam format: 65% passing score. Offered by Salesforce, registered through Webassessor/Kryterion. Updated for Winter '26.
Updated for Winter '26 Release
Last Updated: March 2026 | Exam Version: Winter '26
This exam guide covers the Salesforce AI Associate certification — an entry-level credential validating your understanding of AI fundamentals, Einstein AI capabilities within Salesforce, and responsible AI practices. No prerequisites required. Exam fee: $75.
⚡ What's New in Winter '26
🤖 Agentforce Basics
Understanding autonomous AI agents and how Agentforce fits the Salesforce platform.
🛡️ Responsible AI
Expanded ethical AI content: bias, transparency, accountability and Salesforce's Trusted AI principles.
✨ Einstein Generative AI
Einstein GPT, prompt templates and generative AI use cases across Salesforce clouds.
📊 Exam At a Glance
| Certification Name | Salesforce Certified AI Associate |
| Level | Associate (Entry Level) |
| Prerequisites | None |
| Number of Questions | 40 multiple-choice / multiple-select |
| Duration | 70 minutes |
| Passing Score | 65% |
| Exam Fee | $75 USD |
| Retake Fee | $75 USD (free first retake within 6 months) |
| Delivery | Proctored online or at a testing center |
🎯 Exam Domains & Weightings
1. AI Fundamentals
17%Core AI/ML concepts including supervised, unsupervised and reinforcement learning. Neural networks, natural language processing, and the difference between AI, ML, and deep learning.
🆕 Winter '26: Large language model (LLM) fundamentals and generative AI concepts.
2. AI Capabilities in Salesforce
23%Einstein AI features across Sales Cloud (Einstein Opportunity Scoring), Service Cloud (Einstein Case Classification), and Marketing Cloud (Einstein Send Time Optimization). Einstein GPT and Agentforce platform basics.
🆕 Winter '26: Agentforce autonomous agents and prompt template builder.
3. Ethical Considerations of AI
29%Responsible AI principles: fairness, transparency, accountability, reliability and privacy. Identifying bias in AI models, data privacy requirements, and Salesforce's five Trusted AI principles.
🆕 Winter '26: Ethical deployment guidelines for Agentforce AI agents.
4. AI Strategy and Business Value
31%Identifying AI use cases for business problems, ROI measurement for AI investments, change management for AI adoption, and building an AI strategy roadmap. Understanding when AI adds value vs. when traditional approaches work better.
❓ Sample Exam Questions
A Salesforce administrator is implementing Einstein features across their organization and needs to ensure customer information remains secure and protected from unauthorized access. Which of Salesforce's Trusted AI principles should guide this implementation?
- A. Empowering - enabling users to make better decisions with AI insights
- B. Safe - ensuring data protection, privacy, and security throughout AI operations
- C. Transparent - making AI decision-making processes visible and understandable
- D. Responsible - holding the organization accountable for AI outcomes
A support organization needs to automatically populate case fields such as Subject Category and Urgency Level based on patterns found in their historical case records. Which Einstein AI capability would best accomplish this task?
- A. Einstein Opportunity Scoring to predict case resolution rates
- B. Einstein Case Classification to intelligently assign field values based on case content
- C. Einstein Lead Scoring to identify high-value support accounts
- D. Einstein Send Time Optimization to schedule case notifications
A Salesforce administrator wants to implement Einstein Analytics to forecast opportunity close dates. The solution requires training a model using historical opportunities with known closed dates and associated deal characteristics. Which machine learning approach is most appropriate for this scenario?
- A. Supervised learning with labeled historical opportunity records
- B. Unsupervised learning to discover hidden patterns in unclassified data
- C. Reinforcement learning to optimize prediction accuracy through trial and error
- D. Transfer learning to apply pre-built models from external data sources
A machine learning model deployed in a Salesforce org for resume screening systematically filters out candidates from certain ethnic backgrounds at twice the rate of other applicants, despite no explicit filtering rules being configured. What does this behavior primarily indicate?
- A. The model has been trained on insufficient historical data samples
- B. The model's training data or decision-making process contains systematic prejudice
- C. The model parameters have changed since initial deployment due to new input data
- D. The model is making overly specific distinctions that don't generalize to new populations
A Salesforce administrator needs to configure an Agentforce agent to perform specific tasks like updating customer records and retrieving account information. Which Agentforce component should they configure to enable these capabilities?
- A. Einstein Studio configuration settings
- B. Topics and Actions definitions
- C. Prompt Template customization
- D. Data Cloud Segment mapping
📚 Study Resources
🏃 Trailhead
Complete the AI Associate Certification Prep trail — free, official and covers all exam domains.
Go to Trailhead →📄 Official Exam Guide
Download the official Salesforce AI Associate Exam Guide from Trailhead for exact domain weightings.
Official Guide →💬 Trailblazer Community
Join the AI Associate study group in the Trailblazer Community. Ask questions and share tips with fellow candidates.
Join Community →💡 Top Exam Tips
- Focus on Ethical AI (29%) — this is the highest-weighted domain. Memorise Salesforce's five Trusted AI principles and be able to identify ethical issues in scenarios.
- Business Strategy domain (31%) is also heavy — practise identifying which business problems benefit from AI vs. traditional automation.
- Know Einstein feature names — Case Classification, Opportunity Scoring, Lead Scoring, Send Time Optimization, Next Best Action. Know which cloud each belongs to.
- Learn the difference between AI types — supervised, unsupervised, and reinforcement learning with concrete examples. Common exam question pattern.
- Agentforce basics — understand Topics, Actions, Prompt Templates, and the concept of grounding AI with your org's data.
- Take the Trailhead trail — the official Certification Prep trail aligns directly with the exam. Complete all modules and superbadges before booking.
🙋 Frequently Asked Questions
Ready to Get Certified?
Start with the free Trailhead trail and book your exam when you're scoring 80%+ on practice tests.
Book the Exam →📊 More Data & Analytics Certifications
Explore related Salesforce certifications in the Data & Analytics track