The GC AI chat interface is your primary workspace for interacting with the AI. Here, you can ask legal questions, request document drafting or analysis, get summaries, and leverage GC AI’s various capabilities. Understanding its features will help you work more efficiently.

​
Core Chat Functionality

At its heart, the chat interface works like most messaging applications:

  1. Input Field: Type your requests, questions, or instructions into the text box at the bottom.
  2. Submit: Select the Submit button (or press Enter) to send your message to GC AI.
  3. Conversation History: Your interactions with GC AI are displayed in a chronological thread, making it easy to follow the conversation and refer back to previous exchanges.

​
Key Features Explained

Beyond basic chat, the interface integrates several powerful features:

  • Thinking Indicator: When GC AI is processing your request, especially a complex one involving research or analysis, you’ll see a “Thinking…” indicator. This signifies more than just basic processing; it represents a structured phase where the AI works to understand your query deeply, plan its approach, and execute necessary steps like research or internal reasoning.

    For complex tasks like in-depth legal research, this “Thinking” phase mirrors concepts like the “think” tool developed by Anthropic for their Claude models. It involves:

    • Structured Decomposition: Breaking down your request into manageable sub-problems (e.g., identifying relevant jurisdictions, specific legal questions, necessary statutes or case law).
    • Planning: Formulating a plan, which might include specific search queries for legal databases or web resources. You might see these potential searches appear under the “Thinking” indicator, showing the AI’s planned research steps.
    • Execution & Synthesis: Performing the planned research or analysis steps and then synthesizing the findings into a coherent response.

    This structured thinking process, inspired by advancements in AI reasoning, significantly improves GC AI’s ability to handle multi-step problems accurately and follow instructions or internal policies consistently. Research, like Anthropic’s findings with their “think” tool, shows dramatic performance improvements in complex scenarios, particularly those requiring sequential decision-making and analysis of intricate information (like legal texts or search results). This ensures more reliable and thorough answers for demanding legal research tasks.

  • Research Capability: For questions requiring up-to-date legal information or context beyond its core training data, GC AI can perform targeted research. It might search legal databases or specific web resources to find relevant statutes, case law, or regulatory updates. You may see indicators like “Performing research…” when this happens. This often occurs as part of the “Thinking” phase described above.

  • Web Page Scraping: If you paste a URL (web link) into the chat, GC AI can often access, read, and analyze the content of that web page. This is useful for:

    • Analyzing terms of service or privacy policies directly from a website.
    • Summarizing news articles or blog posts relevant to a legal issue.
    • Checking website copy for compliance.
    • Note: The ability to scrape may depend on the website’s structure and permissions (e.g., pages behind logins might not be accessible).

  • Submit Button: As mentioned, this sends your query to the AI. Ensure your prompt is clear and complete before submitting.

  • Context Window: This refers to the amount of information (text from your current conversation, attached documents, Company Profile details) the AI can “remember” and consider at any given time.

    • GC AI uses advanced models with large context windows, allowing for detailed analysis of lengthy documents and complex conversations. However, there are limits (e.g., a maximum page count for document analysis, currently 150 pages).
    • Be mindful that extremely long conversations might eventually exceed the context window, potentially causing the AI to “forget” details from the very beginning.

  • AI Models Used: GC AI leverages state-of-the-art large language models (LLMs) to provide its capabilities. Currently, a primary model is Anthropic’s Claude 3.7 Sonnet. These models are chosen for their strong reasoning, analysis, and text generation capabilities, particularly in complex domains like law. You can find more information about this model series on the Anthropic documentation site (link to be confirmed and potentially updated).

  • Saved Prompts Integration: You can easily access and use your pre-defined Reusable Prompts directly within the chat interface, saving time on common tasks.

  • Company Profile Context: Information from your Company Profile is automatically considered by GC AI, ensuring responses are tailored to your business context.

​
Tips for Effective Chat Usage

  • Be Specific: The clearer your prompt, the better the response.
  • Provide Context: Upload relevant documents or provide background details.
  • Break Down Complex Tasks: For very complicated requests, consider breaking them into smaller, sequential steps.
  • Review and Refine: Read GC AI’s responses critically. If the output isn’t quite right, refine your prompt and ask again.

Was this page helpful?