HackerRank enables AI assistance for project-based coding questions to mirror real-world development environments. This feature helps candidates experience realistic workflows and allows recruiters to evaluate practical problem-solving more accurately.

The AI Assistant operates in the following modes:

• Unguarded: Allows candidates to interact freely with the AI, similar to real-world AI coding tools and agent-style workflows.

• Guarded (Default): Provides help with syntax, platform navigation, and conceptual guidance but does not generate complete solutions.

The AI Assistant currently supports the following question types:

• Coding

• Database

• Data Science

• Projects

  • Frontend

  • Backend

  • Full-Stack

  • Mobile

  • Generative AI

• Code Repository

Enable AI Assistant for Tests

To enable the AI Assistant:

Step 1: Enable AI Assistant in Company Settings

1. Log in to your HackerRank for Work account using your credentials.

2. Select your profile icon in the upper-right corner.

3. Select Settings from the drop-down menu.

4. Select Test Settings from the left panel under the Company section.

5. In the Test Evaluation section, turn on the AI Assistance for Candidates toggle.

   

6. Click Save Changes.

Step 2: Set up AI Assistant at the section level

1. Go to the Tests tab.

2. Open the test for which you want to add the AI assistant.

3. Hover over the section name and select the Settings icon.

   

4. In the Section Settings dialog, select the AI assistance level you want to enable for candidates:

   

   • No AI assistant

   • Guarded AI assistant

   • Unguarded AI assistant

5. Click Save.

After you enable the AI Assistant at the section level, the section displays an AI Assistant tag.

Candidate experience

When candidates take a test with the AI Assistant enabled in guarded mode, they see AI chat support that allows them to ask for help with syntax, platform navigation, and conceptual guidance.

When candidates take a test with the AI Assistant enabled in unguarded mode, they can interact with it in the following ways:

• Plan Mode: Plan mode lets candidates develop an approach to a problem with the AI Assistant before they write any code. The AI Assistant reads the problem statement and existing code, asks clarifying questions when requirements are ambiguous, and generates a proposed implementation plan in Markdown format for review. Plan mode does not modify code. It only creates a plan for candidates to review before they begin implementation.

  Plan mode also reduces unnecessary rework. Candidates can correct misunderstandings or adjust the proposed approach before the AI generates code. Resolving issues during the planning stage is faster than reviewing and undoing incorrect code after it is written.

• Ask Mode: Candidates can ask questions about the problem statement, code, or specific files. Candidates must tag the problem statement to provide context. The AI Assistant uses the tagged content to generate accurate responses.

  This helps you assess how well a candidate understands the task and how effectively they use the AI Assistant to plan their implementation. It also enables the AI Assistant to apply the same context throughout the interview.

• Agent Mode: Candidates can switch to a mode designed for longer, multi-turn conversations. In this mode, candidates can ask the AI Assistant to write code, edit files, and perform actions directly from prompts. This mode is useful for complex tasks and step-by-step guidance.

  

• Model Switching: Candidates toggle between available models based on their preference or the task requirements. Currently, the AI Assistant supports Claude-sonnet-4.6, Gemini-3.5 Flash, Gemini-3.1 Pro, and GPT-5.4.

  

• Context window: The context window helps candidates understand how much information the AI assistant processes for the current request. For more information, see 📄 Context Window in AI-Assisted Interviews.

Report insights

After the test is complete, the candidate report displays AI-specific interaction data.

• AI Assistant Transcript: This transcript shows the full conversation between the candidate and the AI assistant in the Detailed Report. It is enabled by default for all AI-Assisted Tests.

  

• AI Fluency (optional): Provides recruiters with a holistic view of how candidates interact with the AI assistant during a test. For more information, see 📄 AI Fluency.

Frequently Asked Questions (FAQs)