Ethical AI Content Curation
Drawing inspiration from video generation models advanced generative capabilities (and the potential for deepfakes), this challenge focuses on building an advanced agentic system for ethical AI-generated content curation. Using AutoGen for dynamic agent orchestration and MCP for tool integration, the system will analyze incoming digital content (simulated text, images, or short video descriptions) to determine its AI origin, identify potential misrepresentation, and curate it based on predefined ethical guidelines. This challenge emphasizes the development of MCP-enabled tools that interact with hypothetical AI detection APIs and ethical policy databases. Agents powered by OpenAI o3 (or GPT-5) will collaborate via A2A protocol to assess content authenticity, potential harm, and suggest appropriate labels or moderation actions, showcasing complex decision-making in a sensitive domain.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop an AutoGen multi-agent system with MCP and OpenAI o3 for ethical AI-generated content detection and curation.
How work is evaluated
The evaluation will assess the system's ability to accurately detect AI-generated content characteristics, apply ethical guidelines effectively, demonstrate collaborative decision-making between agents, and provide appropriate curation recommendations.
Challenge datasets
Public datasets, plus the private datasets you have access to.
What you should walk away with
Facilitate secure agent-to-agent (A2A) communication for collaborative content assessment.
Design and implement a dynamic multi-agent system using AutoGen for content analysis.
Develop a system capable of detecting subtle AI-generated content characteristics and suggesting moderation actions.
Utilize OpenAI o3 and/or GPT-5 for advanced content analysis, ethical reasoning, and decision-making.
Integrate MCP-enabled tools for interacting with simulated AI detection and ethical policy APIs.
Reference links and supporting material
Sample data for 2 tasks
Challenge setup
The evaluation will assess the system's ability to accurately detect AI-generated content characteristics, apply ethical guidelines effectively, demonstrate collaborative decision-making between agents, and provide ap...
Challenge input
{'content_type': 'string', 'content_data': 'string', 'context': 'string'}
Agent execution
No agent tool is listed for this challenge.
Evaluated output
{'is_ai_generated': 'boolean', 'ai_confidence_score': 'float', 'deepfake_likelihood': 'float', 'reasoning': 'string', 'detection_tools_used': ['str...
- Ensures AutoGen agents (at least 3 distinct roles) are correctly defined and configured for communication.
- Verifies that at least two MCP-enabled custom tools (simulated or real) are defined and used by agents.
- Detection Accuracy target: 7
- Python execution harness
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
No tool records are attached.
Evaluation contract
- Ensures AutoGen agents (at least 3 distinct roles) are correctly defined and configured for communication.
- Verifies that at least two MCP-enabled custom tools (simulated or real) are defined and used by agents.
Metrics
- Detection_Accuracy · Target 7
- Ethical_Reasoning_Quality · Target 4
- Agent_Collaboration_Score · Target 4
Workflow tasks
- Task 1
Input: {'content_type': 'string', 'content_data': 'string', 'context': 'string'}
Output: {'is_ai_generated': 'boolean', 'ai_confidence_score': 'float', 'deepfake_likelihood': 'float', 'reasoning': 'string', 'detection_tools_used': ['string']}
Execution harness: Python
- Task 2
Input: {'analysis_results': {'is_ai_generated': 'boolean', 'deepfake_likelihood': 'float', 'content_theme': 'string'}, 'ethical_guidelines': ['string']}
Output: {'recommended_action': 'string', 'justification': 'string', 'involved_agents': ['string']}
Execution harness: Python
Recipe freeze marker
No recipe freeze time is recorded.