Normalized claim
Students reached: 330,000 students increase
reached more than 330,000 students
Day of AI Australia / UNSW — multi-agent generative AI simulations with Vertex AI (Win the Farm)
Use case typeAI platformUpdated Jun 13, 2026
Day of AI Australia and partner UNSW Sydney built an immersive educational simulation called Win the Farm to teach AI literacy through a fictional misinformation and narrative game. Students configure AI-powered bots to generate social media posts that influence an election story, while the platform enforces safety and age-appropriate controls at scale.
- Organization
- Day of AI Australia
- Industry
- Education
- Location
- Australia
- Published
- June 2026
Reported outcomes
330,000 students
students reachedOther quantified impact
120,000 postsunique social posts generated350 teamsconcurrent student teams supported
Strategic outcomes
New product / capabilityBuilt a safe generative AI learning simulationRisk & complianceBlocked harmful content in real timeScale & capacityScaled to hundreds of student teamsCost efficiencyReduced manual infrastructure overhead
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Unique social posts generated: 120,000 posts increase
generated more than 120,000 unique social media posts
Normalized claim
Concurrent student teams supported: 350 teams increase
supports 350+ concurrent student teams
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Day of AI Australia
- Provider
- GCP
- Maturity
- Scaled Production
- Linked source
- Google Cloud Customer Stories
Students configure AI-powered bots to generate social media posts that influence an election story, while the platform enforces safety and age-appropriate controls at scale
Customer identity supportedSource describes one deploymentMaturity supported
Primary read
Use case focus
Showing 3 of 4
- 1Generative AI simulation
- 2AI literacy
- 3Multimodal content generation
- Enable 12–16-year-old students to interact with generative AI safely in a browser-based game.
- Block harmful, age-inappropriate, or policy-violating content in real time.
- Scale to hundreds of concurrent student teams with privacy-preserving infrastructure and no student PII storage.
- Built a multi-agent simulation platform on Google Cloud.
- Used Vertex AI for inference requests, Model Garden to choose and deploy models, Gemini in Vertex AI to generate content and images, and Shield Gemma to evaluate output before it reaches students.
- Used Compute Engine for orchestration and Cloud SQL for encrypted game state and teacher logs without storing student PII.
- Generated 120,000+ unique social posts.
- Supported 350+ concurrent student teams.
- Reached more than 330,000 students.
- Reportedly had no safety incidents.
- Reduced manual infrastructure overhead for a lean team.
Architecture
The platform uses Compute Engine VMs to orchestrate background tasks for 350+ concurrent student teams. Every 10 minutes, the system processes bot interactions using Vertex AI inference requests. Open Gemma models generate social posts, Shield Gemma evaluates them for policy compliance before they reach the student-facing Barn Wall feed, and Cloud SQL stores encrypted game state and teacher logs without student PII. Model Garden on Vertex AI is used to select and deploy the model mix, and Gemini in Vertex AI pre-generates images and NPC backstories.
Implementation partners1
Sources & evidence1
Evidence: Medium65/100Evidence strength
- Customer explicitly identified
- Deployment status explicitly supported
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Jun 7, 2026Publisher: Google CloudEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
No published comments yet.