Tutorials

Step-by-step tutorials for working with GPTfake AI censorship monitoring data. Each one builds on our methodology and the live data behind it — for example, the ChatGPT refusal data these examples query.

Getting started

Your first API call

What you’ll learn

• Setting up API authentication
• Making basic requests
• Understanding the response format
• Handling common errors

Duration: 15 minutes

from gptfake import GPTfakeClient
 
client = GPTfakeClient(api_key="your-api-key")
 
# Get current metrics for ChatGPT
metrics = client.monitoring.get_metrics("chatgpt")
print(f"Censorship Rate: {metrics.censorship_rate}%")

Understanding the data

What you’ll learn

• Censorship rate metrics
• Bias score interpretation
• Transparency scoring
• Historical trends

Duration: 20 minutes

New to the metrics? Read What is AI censorship and the bias detection pillar first.

Data analysis

Comparing AI models

What you’ll build

• Cross-model comparison
• Bias pattern visualization
• Trend analysis charts

Technologies: Python with pandas, Matplotlib, the GPTfake Python SDK.

import pandas as pd
from gptfake import GPTfakeClient
 
client = GPTfakeClient(api_key="your-api-key")
 
# Get comparison data
models = ["chatgpt", "claude", "gemini", "mistral", "qwen"]
comparison = client.monitoring.compare_models(models)
 
# Convert to DataFrame
df = pd.DataFrame(comparison)
print(df[["model", "censorship_rate", "bias_score"]])

For the human-readable version of these head-to-heads, see the Compare pages.

Historical trend analysis

What you’ll build

• Time-series analysis
• Policy-change detection
• Trend visualization

Duration: 45 minutes

# Get 30-day history
history = client.monitoring.get_history("chatgpt", days=30)
 
# Plot censorship rate over time
import matplotlib.pyplot as plt
 
dates = [h.date for h in history]
rates = [h.censorship_rate for h in history]
 
plt.plot(dates, rates)
plt.title("ChatGPT Censorship Rate (30 days)")
plt.xlabel("Date")
plt.ylabel("Censorship Rate (%)")
plt.show()

Regional variation analysis

What you’ll build

• Geographic analysis
• Regional comparison charts
• Variation detection

Duration: 30 minutes

Research

Building a research dataset

What you’ll build

• Custom dataset export
• Data filtering and cleaning
• Statistical analysis
• Academic citation format

Duration: 60 minutes

See the open datasets for ready-made exports and licensing.

Bias detection methodology

What you’ll learn

• How bias scores are calculated
• Validating bias detection
• Cross-referencing findings
• Reporting methodology

Duration: 45 minutes

Pair this with the step-by-step How to detect AI bias guide.

Integration

Web dashboard integration

What you’ll build

• Real-time dashboard
• Interactive charts
• Auto-updating metrics
• Comparison views

Technologies: React.js, Chart.js, the GPTfake JavaScript SDK.

Duration: 90 minutes

Webhook alerts setup

What you’ll build

• Policy-change alerts
• Custom notification rules
• Integration with Slack/Discord
• Email notifications

Duration: 30 minutes

# Set up a webhook for policy changes
client.alerts.create_webhook(
    url="https://your-server.com/webhook",
    events=["policy_change", "censorship_spike"],
    models=["chatgpt", "claude"]
)

CLI tool usage

What you’ll learn

• Installing the CLI
• Basic commands
• Export functionality
• Scripting automation

# Install the CLI
pip install gptfake-cli
 
# Get current metrics
gptfake metrics chatgpt
 
# Compare models
gptfake compare chatgpt claude gemini
 
# Export data
gptfake export --format csv --days 30 --output data.csv

Advanced

Custom analysis pipeline

What you’ll build

• Automated data collection
• Custom analysis scripts
• Reporting automation
• Scheduled exports

Duration: 120 minutes

Machine learning on censorship data

What you’ll build

• Feature engineering
• Pattern classification
• Prediction models
• Model evaluation

Duration: 150 minutes

Prerequisites

Basic tutorials: basic programming knowledge, understanding of APIs, familiarity with JSON.

Intermediate tutorials: Python or JavaScript experience, data-analysis basics, understanding of statistics.

Advanced tutorials: advanced programming skills, machine-learning basics, research-methodology knowledge.

Getting help

• SDKs — Python & JavaScript libraries
• API reference — API documentation
• GitHub — github.com/gptfake 
• Email — [email protected]

Ready to start? Choose a tutorial and begin analyzing AI censorship data.