Minister Roban On Use Of AI In Department
Speaking in the House of Assembly today [Nov 15], Minister of Home Affairs Walter Roban gave an update on the integration of artificial intelligence and machine learning within the Land Title and Registration Department.
Minister Roban said, “This significant step forward reflects our commitment to modernizing land registry processes, enhancing efficiency, and building a more sustainable future for the Department.
“This enables individuals to prove ownership, prevents disputes, and promotes legal certainty in property transactions across the island. Mr. Speaker, since 2020, the Department has faced an accumulating backlog of approximately 800 first registration cases.
“This increase was due to the impact of the COVID-19 pandemic, the surge in property sales, and an increase in voluntary registrations. These factors highlight the need for an efficient and secure method of processing title registrations.”
The Minister’s full statement follows below:
Mr. Speaker, Honourable Members of the House, and people of Bermuda, I rise today to present an update on the integration of artificial intelligence [AI] and machine learning [ML] within the Land Title and Registration Department [LTR]. This significant step forward reflects our commitment to modernizing land registry processes, enhancing efficiency, and building a more sustainable future for the Department. Mr. Speaker, the purpose of this initiative is to harness the transformative potential of AI and ML to address the backlog in land title cases.
“This integration not only meets an immediate operational need but also establishes a sustainable solution for the long-term management of Bermuda’s land registry. Mr. Speaker, the Land Title and Registration Department was established in 2006 with the core responsibility of ensuring secure property ownership by maintaining official land title records.
“This enables individuals to prove ownership, prevents disputes, and promotes legal certainty in property transactions across the island. Mr. Speaker, since 2020, the Department has faced an accumulating backlog of approximately 800 first registration cases. This increase was due to the impact of the COVID-19 pandemic, the surge in property sales, and an increase in voluntary registrations. These factors highlight the need for an efficient and secure method of processing title registrations.
Mr. Speaker, our current resources have proven insufficient to address this increased volume of work. Although we have embarked on a project to scan deeds, this step alone is not enough to clear the backlog, as each document still requires a detailed review, verification, and entry process, which is both time-consuming and resource intensive. Mr. Speaker, this backlog impacts not only our department but also other critical services.
It delays the flow of land tax information to the Office of the Tax Commissioner, affects realtors, the Land Valuation Department, mortgage banks, and has broader implications for our economy. Mr. Speaker, we have carefully considered several options to address this issue. Among these were offering overtime to existing staff, hiring additional staff at an estimated cost of $470,000, outsourcing to overseas vendors, using trainees, and engaging retired legal professionals. Each option had significant drawbacks, including high costs, risk to indemnity, and resource constraints.
After a thorough review, the Ministry of Home Affairs determined that an AI-driven solution, like the model implemented by the UK Land Registry, is the most sustainable and cost-effective option. This approach promises to meet current needs while preparing the Department for future demands. Mr. Speaker, We are proceeding with a partnership with FluentData as our AI vendor.
After successful testing, the AI system demonstrated that it could extract required information from deeds in seconds—a task that would otherwise take human officers several hours. This technology will streamline document processing, reduce caseworker workload, and integrate seamlessly into existing workflows.
Mr. Speaker, the implementation plan consists of several key steps. First, the backlog of 800 case files has already been scanned, organized, and returned to the Department. Second, we have secured approval for a subscription-based Google AI service to process these digital files. We anticipate that the data extraction process will be completed within 12 weeks or sooner. Once the data has been extracted, it will be imported into Landfolio, our registry system.
The LTR staff will verify data accuracy before integrating it into the official records. We also anticipate minimal ongoing maintenance through a consumption-based subscription model, ensuring updates are efficient and cost-effective. Mr. Speaker, the initial cost for AI implementation is expected to be approximately $50,000, with minimal ongoing expenses. This is a far more economical solution than hiring additional personnel, providing a scalable model that addresses both immediate and future needs.
The integration of AI and ML within the Land Title and Registration Department is a progressive move for Bermuda. By leveraging this technology, we ensure a high level of accuracy, security, and efficiency, while minimizing risks to government indemnity and improving service delivery.
Mr. Speaker, the rollout of the LTR AI solution is to take place by December 20th, 2024 and the completion date anticipated to be the end of January 2025. Mr. Speaker, I must apologize to the general public for any inconvenience caused by these issues. We are confident that this initiative, supported by the Department, and working alongside the Digital Transformation Team, will help us eliminate the backlog, facilitate ongoing operations, and position the Department to handle future workload demands.
Thank you, Mr. Speaker.
Read More About
Category: All, technology
