AI you can trust with real estate decisions.
We build and deploy AI to help our clients make better property and location decisions — safely, securely and transparently. Privacy-first, human-accountable, and grounded in recognized governance frameworks.
AI that enhances insight — never at the expense of privacy or trust.
We integrate AI to improve accuracy, efficiency and insight across our data platform, products and services. Our approach centres on protecting data, keeping people accountable, and maintaining clear governance everywhere AI is used.
In short
- Client data is never used to train AI models that are shared with other clients.
- Cloud AI runs with tenant-level isolation — your information is never used to train or enhance third-party models.
- We apply industry-standard controls: encryption, de-identification, least-privilege access, data minimization and clear retention and deletion.
- Our Chief Technology Officer is accountable for AI risk, supported by an AI Governance Group and internal risk assessments.
- Humans stay accountable for every final decision — AI only supports expert judgment.
Five principles behind every AI decision.
Transparency & accountability
We're clear about where and how AI touches property and location data, and people — not models — remain responsible for the decisions it informs.
Governance & risk management
AI operates inside a defined governance framework, with policies, oversight and risk assessment across its full lifecycle.
Fairness & bias management
We evaluate our own models and scrutinize third-party providers for fairness, so location and demographic AI doesn't entrench bias.
Human-centred by design
We acquire, design and deploy AI to augment expert judgment — analysts, agents, underwriters — and to respect human oversight at all times.
Safe, purpose-bound deployment
Every AI feature ships with safeguards and stays within the limits of what it was built to do, prioritizing safety, privacy and data protection.
Where AI shows up at Homicity.
Internal productivity
AI helps our teams draft, summarize, research and automate internal workflows. These tools run only on internal information or data an employee provides for their own task — never client data.
Product AI features
Our platforms include AI-assisted search, summarization and generative assistance. Each feature ships with built-in guardrails and pre-release evaluations for accuracy, safety and responsible use.
Data development
We build the Neighbourly data platform from public, licensed and partner sources — never client data — using statistical modelling and machine learning to estimate neighbourhood-level attributes across Canada.
Advisory & analytics services
Our analysts apply AI to support modelling, forecasting and insight on client-provided data, under project-specific controls and contractual restrictions.
How we keep client and consumer data safe.
Secure processing
Cloud-based AI runs on enterprise services with tenant-level isolation and strong controls, so client information stays private and is never used to train or enhance third-party models.
Security by design
Encryption, least-privilege access management, data minimization, monitoring and audit logging are standard across our platform and services.
Canadian compliance
We align with PIPEDA and applicable provincial privacy law, and host data in Canadian data centres.
Data lifecycle
Retention follows classification-based practices, is governed by client data-sharing agreements, and supports secure deletion on request.
Safeguards at every layer.
Isolation
AI features run in cloud or private environments with tenant- or instance-level isolation and strong data-handling controls.
Inference-only by default
Client data is not used to train or improve AI models for other customers or third parties.
Input protections
Prompt-injection defenses, input validation and data-minimization controls reduce the risk of unintended instructions or misuse.
Output constraints
System prompts, allowed-content policies and domain-specific instructions keep AI responses relevant and supported.
Abuse & anomaly monitoring
Rate limiting, anomaly detection and audit logging deter misuse and support investigation.
Quality checks
AI features undergo pre-release evaluations for accuracy, safety and reliability, and we monitor and adjust controls over time.
Clear ownership, all the way to a person.
- Our AI Governance Group oversees AI strategy, risk, policy and lifecycle management.
- The Chief Technology Officer is accountable for responsible AI and overall AI risk management.
- Internal AI risk assessments guide safeguards, evaluations and deployment decisions.
- Humans remain accountable for final decisions across all AI-enabled workflows.
Built on recognized standards.
Our responsible-AI program is anchored in the NIST AI Risk Management Framework and aligns with the OECD AI Principles and Canada's Voluntary Code of Conduct on Advanced Generative AI Systems.
Responsible AI — common questions.
Does Homicity use my data to train AI?
No. Client data is never used to train AI models that are shared with other clients, and cloud AI runs with tenant-level isolation so your information is never used to train or enhance third-party models.
What frameworks guide Homicity's responsible AI program?
Our program is anchored in the NIST AI Risk Management Framework and aligns with the OECD AI Principles and Canada's Voluntary Code of Conduct on Advanced Generative AI Systems.
Does Homicity let AI make decisions automatically?
No. AI is used to support expert judgment. Humans remain accountable for final decisions across every AI-enabled workflow.
Is client data involved in how Homicity builds its data products?
No. The Neighbourly data platform is built from public, licensed and partner sources — never client data — using statistical modelling and machine learning to estimate neighbourhood-level attributes across Canada.
Where is Homicity's data stored?
In Canadian data centres, aligned with PIPEDA and applicable provincial privacy law, with encryption and strict access controls by default.
How does Homicity keep its AI features safe?
Guardrails include tenant isolation, prompt-injection defenses, output constraints, abuse and anomaly monitoring, audit logging, and pre-release evaluations for accuracy and reliability.
Questions about how we use AI?
Talk to our team about our responsible-AI program, data handling and governance.