DTTP AI PM

Veritarte

Veritarte is an AI-Generated Content Labeling and Filtering tool that aims to address the growing need for transparency in Etsy’s digital art marketplaces. It automatically detects, labels, and filters AI-generated artwork. Veritarte seeks to enhance user trust, improve buyer experience, and support human artists.

Problem Statement

Research findings reveal that 55% of digital art shoppers on Etsy find it difficult to distinguish between AI-generated artwork and human-created artwork. This makes the experience of shoppers that value human-created digital art frustrating as they have committed their money and time to the process. This undermines trust in Etsy’s marketplace and hinders informed purchasing decisions. 

Although some users do not mind AI-generated artwork, most users would like to still make informed decisions and consciously choose AI-generated digital art, if they are to make that decision. 67% of users prefer human-created digital art and 89% are willing to pay a premium for that category, revealing an opportunity area for sellers that specialize in human-created digital art. 

By implementing an AI-driven solution for content labeling and filtering, Etsy can enhance transparency, empower buyers to make informed choices, and support sellers in accurately showcasing their creations. 

Problem statement: How might we enable clear labeling and filtering of AI-generated content so that buyers can easily distinguish between AI-generated and human-created art and make informed purchasing decisions?

Goals

  • Measure transparency impact by tracking key metrics such as:
    • Number of page visits to AI-generated vs. human-created art
    • Time spent on labeled AI-generated vs. human-created art pages
    • Bounce rates for AI vs. non-AI listings
  • Development of AI models to detect AI-generated artwork.
  • Implementation of labeling for AI-generated content.
  • Track purchasing behavior to assess marketplace impact: 
    • Number of AI-generated art purchases
    • Number of human0created art purchases
    • Conversion rates for AI-generated vs. Human-created artwork 
  • Introduction of filtering options for buyers to choose between AI-generated and human-created art.
  • Enhance transparency in the digital art marketplace.
  • Build trust with buyers by providing clear labeling of AI-generated art and monitoring:
    • Percentage of users who proceed to purchase after viewing verification labels
    • User feedback ratings on AI vs. human-created art
  • Support human artists by allowing buyers to easily filter and find human-created artwork.

Non-Goals

  • Verification of physical products.
  • Monitoring non-digital art categories.

User Stories and Use Cases

 User Story 1: Emily – The Art Enthusiast 

As Emily, an art enthusiast, I want to be able to see clear labels on AI-generated art so that I can support human artists and make informed purchasing decisions.

 Use Case 1: AI-Generated Content Labeling 

Develop an AI model that detects AI-generated artwork and applies clear labels. 

User Story 2: Lisa – The Casual Buyer 

As Lisa, a casual buyer, I want to filter my search results to show only human-created art so that I can find affordable and authentic digital artwork.

 Use Case 2: AI-Powered Search Filtering 

Introduce a filtering system that allows users to choose between AI-generated and human-created art.

 User Story 3: Sellers on Etsy 

As a digital art seller on Etsy,  I want my human-created art to be verified and labeled as authentic so that buyers can trust my work.

Enable an “Authenticity Verified” badge for human artists using AI-based verification methods.

Customers and Business Impact

Customers Impact

Veritarte will increase trust with buyers of digital art on Etsy. They will trust in the aunthenticity of the artwork they purchase, knowing Veritarte will have help them to easily distinguish between AI-generated and human-created art. Transparency in labeling will reduce buyers feeling deceived, leading to higher satisfaction. The user interface enhancements, such as visual indicators and customizable search preferences, will create a more intuitive and enjoyable shopping experience. 

Business Impact

For Etsy, Veritarte will have the following impact:

  • Increased Customer Loyalty & Retention: By fostering trust and transparency, Etsy aims to achieve a 10–15% increase in repeat purchase rates and improve customer lifetime value (CLV) by 12% within the first year.
  • Market Differentiation: Veritarte positions Etsy as a leader in ethical digital art commerce, aiming for a 20% increase in new user acquisition from artists and buyers seeking transparent marketplaces.
  • Higher Sales & Revenue: Improved buyer confidence will lead to:
    • A 15% increase in conversion rates for human-created and AI-generated art.
    • A 10% rise in average order value (AOV) as customers feel more secure in their purchases.
    • A 25% boost in engagement metrics (e.g., time spent on listings, and searches using filtering options).
  • Enhanced Brand Reputation: By committing to transparency and actively supporting human artists, Etsy expects:
    • A 20% increase in positive customer reviews and ratings.
    • A 30% rise in social media mentions and word-of-mouth referrals from satisfied buyers and artists.

Solutions

Alternative Solutions

Manual Verification and Labeling:

This solution involves a team of human reviewers manually verifying and labeling each piece of artwork on Etsy's platform. Reviewers would examine the artwork to determine whether it is AI-generated or human-created and then apply the appropriate labels.

Why it's not optimal:

  • Scalability: As the volume of digital artwork grows, it would become increasingly challenging to manually verify and label each piece, leading to inefficiencies and potential bottlenecks.
  • Consistency: Human reviewers might have different interpretations of what constitutes AI-generated art, resulting in inconsistent labeling and user confusion.
  • Cost: The labor-intensive nature of this solution would incur significant operational costs, making it less economically viable compared to an automated solution.

Seller Self-Reporting:

In this approach, sellers would be responsible for self-reporting whether their artwork is AI-generated or human-created. Sellers would complete a declaration during the product listing process, indicating the nature of their artwork.

Why it's not optimal:

  • Accuracy: Self-reporting relies on the honesty and accuracy of sellers, which could lead to mislabeling or dishonest practices, undermining user trust.
  • Lack of Verification: Without a verification mechanism, buyers may still question the authenticity of the artwork, resulting in persistent transparency issues.
  • Complexity: This approach places an additional burden on sellers, potentially deterring some from listing their artwork on Etsy or leading to incomplete or inaccurate declarations.

User Flow

Please find the link to the flowcharts for Experiences are HERE

Experience 1: Discovering AI Labeling Feature

User: Digital art buyer browsing Etsy for new artwork

  1. Landing Page:
  • Users arrive at Etsy’s digital art section.
  • Notifications inform users about the new AI-powered labeling feature.
  1. Search and Browse:
  • Users search for digital art using keywords.
  • A pop-up explains the labeling feature and its benefits.
  1. Artwork Listings:
  • Search results display clear labels for AI-generated or human-created art.
  • Visual icons accompany the labels for quick identification.
  1. Product Page:
  • Click on an artwork to view details.
  • The page shows the label and a verification badge if it’s human-created.
  1. Post-Purchase:
  • After purchasing, users receive a follow-up email thanking them for their feedback.

Outcome: The user gains confidence in identifying and purchasing their preferred type of artwork

Experience 2: Using Filtering Options

User: Art-enthusiast particular about human created artwork

  1. Landing Page:
  • The user searches for “digital watercolor prints” and filters results using the "Human-Created" filter powered by Veritarte.
  • Listings update instantly to only show verified human-created artwork, complete with an authenticity badge
  1. Activating Filters:
  • Users access filter settings under the search bar.
  • Options include:
  • "Human-Created Art"
  • "AI-Generated Art"
  1. Filtered Search Results
  • Search results update based on selected filters.
  • Users see clear labels and relevant artworks.
  1. Product Page:
  • Click on filtered artwork to view more details.
  • Verification badges appear if the art is human created.
  • Users provide feedback on filtering accuracy.
  1. Checkout:
  • Users proceed with purchasing filtered artwork with added trust and confidence.

Outcome: The user finds and purchases a verified human-created piece, assured that they are supporting traditional artists.

Experience 3: Checking AI vs. Human-Created Labels on a Product Page

User: A casual Etsy shopper curious about AI-generated art.

  1. Product Page:
  • The user clicks on a digital print and notices a label stating "AI-Generated."
  • A tooltip explains how Veritarte identifies AI-generated art.
  1. Exploring Similar Listings:
  • The user clicks on related AI-generated artworks.
  • They compare AI and human-created options.
  1. Decision Making:
  • The user selects their preferred artwork based on transparency labels.
  • They proceed to checkout or continue browsing.

Outcome: The user better understands AI’s role in digital art and engages with the transparency features.

Experience 4: Post-Purchase Experience & Feedback Collection

User: Buyer who just purchased a digital artwork.

  1. Order Confirmation & Label Reminder:
  • After purchasing, the user sees a confirmation page summarizing their order.
  • A message highlights the AI/human label of their purchased artwork.
  1. Follow-Up Email:
  • The user receives a thank-you email from Etsy.
  • The email includes a prompt to provide feedback on Veritarte’s labeling feature.
  1. Survey Participation:
  • The email links to a short survey about:
  • The usefulness of the labels.
  • Their trust level in AI transparency.
  • Suggestions for improvements.
  1. User Engagement & Future Improvements:
  • Users who complete the survey may receive a discount or loyalty incentive.
  • Feedback is analyzed to refine Veritarte’s labeling accuracy and usability.

Outcome: Etsy collects valuable user insights to refine Veritarte and enhance user trust in AI-powered transparency.

Wireframe

Veritarte search filter

A paper with writing on itAI-generated content may be incorrect.
A drawing of a house and a price listAI-generated content may be incorrect.

Veritarte automatically detects and labels AI-generated art

  Functional Requirements

Requirement Priority

1. Detect AI-generated artwork with an accuracy rate of at least 90%.

2. Apply clear and consistent labels to AI-generated content.

3. Provide filtering options in the search interface to distinguish between AI-generated and human-created art

4. Enable an “Authenticity Verified” badge for human-created artwork.

5. Provide real-time updates to labels and filtering options as new digital artwork is added to the platform

6. Integrate seamlessly with Etsy’s existing search algorithm to ensure a smooth user experience

7.Allow users to customize their search preferences to show or hide Ai-generated artwork

8.Provide visual indicators on search results and product pages to clearly distinguish AI-generated art from human created art

9.Include an option for sellers to manually verify and label their artwork as human-created, subject to Veritarte verification

10.Support multiple languages to cater to Etsy’s diverse global user base

11. Allow buyers and sellers to leave feedback on the accuracy of the AI-generated content labels, which can be used to improve the model over time.

Non-Functional Requirements

  • The system should process labeling and filtering within 1-2 seconds to ensure a seamless user experience.
  • The system should be scalable to handle the growing volume of digital artwork on Etsy’s platform.
  • The system should ensure data privacy and security, complying with relevant regulations and standards.

Integration of Veritarte into Etsy’s Digital Art Marketplace

Veritarte seamlessly integrates into Etsy’s digital art marketplace by providing AI-powered transparency features that help buyers distinguish between human-created and AI-generated artwork. The integration enhances trust, improves the shopping experience, and supports artists who seek authenticity verification.

Key aspects of Veritarte’s integration include:

  • AI Labeling System: Veritarte automatically scans artwork listings and applies an "AI-Generated" or "Human-Created" label to each piece, ensuring buyers can make informed decisions.
  • Verification for Sellers: Artists can verify their work through an authentication process that grants a "Human-Created" badge, adding credibility to their listings.
  • Filtering and Search Enhancements: Buyers can filter search results based on artwork origin, ensuring they see only human-created or AI-generated works based on preference.
  • Product Page Enhancements: Each artwork listing displays a label and an explanation tooltip about how the categorization works, with a feedback section for users to confirm accuracy.
  • Post-Purchase Engagement: Buyers receive follow-up emails encouraging feedback on the labeling system, allowing Etsy to improve the feature based on real user input.

AI Model & Data Strategy

Veritarte leverages a hybrid AI approach combining deep learning image recognition models, metadata analysis, and user feedback loops to classify artwork accurately. The AI model will be trained using:

  • Supervised Learning on Labeled Datasets – Training data will include known AI-generated artworks from various sources (e.g., MidJourney, Stable Diffusion, DALL-E) and verified human-created pieces from established artists.
  • Feature Extraction – The AI will analyze image structures, brushstrokes, metadata, and inconsistencies to detect AI-generated patterns.
  • Metadata & Provenance Tracking – Veritarte will scan file metadata and generation history where available to validate authenticity.

Handling False Positives/Negatives:

  • Users can flag misclassified artworks through a feedback system.
  • A confidence threshold will determine when human review is required.
  • Continuous Model Training – The AI will retrain periodically using Etsy community feedback, refining classification accuracy.

FAQ’s

Frequently Asked Questions (FAQ)

1. What is the purpose of the AI-Generated Content Labeling and Filtering solution? The purpose of this solution is to enhance transparency in Etsy's digital art marketplace by using AI to detect, label, and filter AI-generated artwork. This allows buyers to make informed purchasing decisions and supports human artists by clearly distinguishing their work.

2. How will the AI technology be integrated into Etsy's platform? The AI technology will be integrated seamlessly into Etsy's existing infrastructure, including its search algorithm. Users will be able to see clear labels on AI-generated artwork and use filtering options to distinguish between AI-generated and human-created art.

3. What are the benefits of this solution for buyers? Buyers will benefit from increased trust and confidence in the authenticity of the artwork they purchase. The solution provides clear labels and filtering options, enabling buyers to support human artists and make well-informed purchasing decisions.

4. How will this solution support human artists? Human artists will benefit from an "Authenticity Verified" badge, which distinguishes their work from AI-generated art. This helps human artists stand out in a competitive market and attract buyers who value genuine, handmade artwork.

5. What measures are in place to ensure the accuracy of AI-generated content labeling? The AI model will be trained using a comprehensive dataset of AI-generated and human-created artwork to achieve a high accuracy rate of at least 90%. Regular updates and training will be conducted to maintain and improve accuracy over time.

6. How will user feedback be incorporated into the system? Buyers will have the option to leave feedback on the accuracy of the AI-generated content labels. This feedback will be used to continuously refine and improve the AI model, ensuring it meets user expectations.

7. What impact will this solution have on Etsy's business? The solution is expected to increase customer loyalty and retention, differentiate Etsy from other marketplaces, and enhance brand reputation. It will also drive higher sales and revenue by improving user experience and transparency.

8. Are there any non-AI alternatives considered for this solution? Yes, non-AI alternatives such as manual verification and labeling, and seller self-reporting were considered. However, these alternatives present challenges in scalability, consistency, and accuracy, making them less effective compared to an AI-driven solution.

9. How will the system handle a large volume of digital artwork? The system will be designed to handle a large volume of digital artwork without significant performance degradation. It will be scalable to accommodate the growing number of artworks on Etsy's platform.

10. What are the success metrics for this solution? Success metrics include increased user trust and satisfaction, accurate identification of AI-generated art, improved purchase decisions, reduced customer complaints, higher sales of human-created art, and enhanced operational efficiency.

Success Metrics

Outcome

Measure

Current | Target

Priority

1. Increased user trust Customer feedback and satisfaction scores

Current: N/A |Target: 90%

2. Accurate identification of AI art 90% accuracy of reporting on AI-generated artwork

Current: N/A |Target: 90%

3. Enhanced user engagement Increase in user interactions with digital art

Current: 60%|Target: 80%

4. Reduced customer complaints, Number of complaints related to undisclosed AI art

Current: 60%|Target 90%

5. Higher sales of human-created art. Increase in sales metrics for human-created art

Current: 40%|Target 70%

6.User satisfaction with filtering, User satisfaction survey results

Current: 55%|Target 85%

7. Operational efficiency , Reduction in time to label/filter artwork

Current: N/A | Target: 1 min

 Timelines and Milestones

Item

Timeline

Exit criteria

Research and development

3 months

Research completed, Development plan finalized, required resources allocated

AI Model Training and Testing

2 months

AI model trained with a minimum 90% accuracy rate, Successful internal testing and validation

Implementation and Integration

2 months

AI model integrated with Etsy’s platform, Seamless integration with existing search algorithm

User testing and feedback

1 month

User testing conducted, Feedback collected and analyzed, Necessary adjustments made based on feedback

Launch monitoring

1 month

Successful deployment, Monitoring mechanisms in place, Initial performance metrics reviewed and analyzed. 

Appendix

User Flow for Veritarte

Learnings

Product Manager Learnings:

Motunrayo Babatunde

Designer Learnings:

Designer Learnings:

Jo Sturdivant

  1. Adapting to an Established Team: Joining the team in week 6 of 8 was challenging, as I had to quickly adapt to existing workflows, dynamics, and goals. This mirrors real-world situations where you often integrate into teams mid-project, and flexibility is essential.
  2. Work-Blocking for Efficiency: With only two weeks to complete the project, I learned the importance of a structured work-blocking system. This approach allowed me to manage my time effectively and meet deadlines under pressure.
  3. Making Data-Driven Design Decisions: Unlike my past projects, I had to rely on research conducted by others. This was a valuable experience in using pre-existing data to guide design decisions, helping me focus on the core insights without starting from scratch.

Developer Learnings:

Developer Learnings:

Vanady Beard

&

As the back-end developer, I learned how important it is to create efficient and reliable systems that support the entire application. This experience also taught me the importance of optimising the database and ensuring the backend is scalable and easy to maintain.

Developer Learnings:

Stephen Asiedu

&

As a back-end developer, I've come to understand the importance of being familiar with various database systems and modules. This knowledge enables me to build diverse applications and maintain versatility in my work. I've also learned that the responsibility for making the right choices rests on my shoulders, guided by my best judgement.

Developer Learnings:

&

Developer Learnings:

Maurquise Williams

&

  1. Process of Creating an MVP: Developing a Minimum Viable Product (MVP) taught me how to focus on delivering core functionalities balancing between essential features and avoiding scope creep.
  2. Collaboration in a Real-World Tech Setting: This experience taught me how to collaborate efficiently in a fast-paced tech environment, keeping the team aligned and productive, even while working remotely across time zones.
  3. Sharpening Critical Thinking and Problem-Solving Skills: This experience honed my ability to think critically and solve problems efficiently. By tackling challenges and finding quick solutions, I sharpened my decision-making and troubleshooting skills in a dynamic, real-world setting.

Developer Learnings:

Jeremiah Williams

&

All in all this experience was very awesome I learned that in coding with others being transparent is key

Developers Learnings:

Justin Farley

&

I learned how important communication is when working with a team. Communication provides understanding, advice, ideas, and much more. While working with the product team, I’ve found that communication keeps everything flowing smoothly. Working with a team also showed me that every member brings something different to the table and we all have to work together in order to align and meet our end goal.

Full Team Learning