Multi-Agent System for AI-Generated Content Verification & Compliance
Inspired by the 'Human Authored' logo initiative and growing concerns about AI-generated content, this challenge requires building a sophisticated multi-agent system using LangChain (specifically LangGraph for orchestration). The system will analyze content for authenticity, detect potential AI generation, and check for compliance against ethical guidelines. Utilizing Gemini 3 Flash for rapid analysis and summarization, the agent team will coordinate using graph-based workflows. Cognee will provide long-term memory for learning content patterns and historical decisions. Giskard will be integrated for continuous evaluation, bias detection, and governance, ensuring the system remains ethical and performs reliably. Coplay AI will serve as an interactive interface for users to submit content and receive detailed explanations of the analysis.
What you are building
The core problem, expected build, and operating context for this challenge.
Inspired by the 'Human Authored' logo initiative and growing concerns about AI-generated content, this challenge requires building a sophisticated multi-agent system using LangChain (specifically LangGraph for orchestration). The system will analyze content for authenticity, detect potential AI generation, and check for compliance against ethical guidelines. Utilizing Gemini 3 Flash for rapid analysis and summarization, the agent team will coordinate using graph-based workflows. Cognee will provide long-term memory for learning content patterns and historical decisions. Giskard will be integrated for continuous evaluation, bias detection, and governance, ensuring the system remains ethical and performs reliably. Coplay AI will serve as an interactive interface for users to submit content and receive detailed explanations of the analysis.
Shared data for this challenge
Review public datasets and any private uploads tied to your build.
How submissions are scored
These dimensions define what the evaluator checks, how much each dimension matters, and which criteria separate a passable run from a strong one.
AllAgentsOperational
All defined agents in the LangGraph workflow must execute without critical errors.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
GiskardTestPassRate
A minimum percentage of Giskard bias and robustness tests must pass.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
AIGenerationF1Score
F1 score for classifying AI vs. human content, reflecting both precision and recall. • target: 0.88 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
ExplanationClarityScore
Subjective or automated score (e.g., readability) for the clarity of explanations provided by Coplay AI. • target: 4 • range: 1-5
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Master LangGraph for designing complex, stateful multi-agent workflows with conditional routing and human-in-the-loop steps.
Implement a specialized 'Content Analyzer' agent powered by Gemini 3 Flash for real-time detection of AI-generated text patterns and stylistic anomalies.
Design a 'Compliance Agent' that uses Gemini 3 Flash to cross-reference content against predefined ethical and regulatory guidelines, outputting a detailed compliance report.
Integrate Cognee to provide agents with a shared, persistent memory layer for learning from past content analyses, improving accuracy over time, and reducing redundant processing.
Leverage Giskard for setting up an evaluation harness to continuously monitor the agents' performance, detect biases in content assessments, and ensure model robustness against adversarial inputs.
Build a user-facing interface using Coplay AI that allows submission of content and provides clear, interactive explanations of the multi-agent system's analysis and conclusions.
Develop custom tools within LangChain for agents to interact with a simulated content database and Giskard's testing platform.
[ok] Wrote CHALLENGE.md
[ok] Wrote .versalist.json
[ok] Wrote eval/examples.json
Requires VERSALIST_API_KEY. Works with any MCP-aware editor.
DocsAI Research & Mentorship
Participation status
You haven't started this challenge yet
Operating window
Key dates and the organization behind this challenge.
Find another challenge
Jump to a random challenge when you want a fresh benchmark or a different problem space.