Building Legal AI for Public Service Across Borders
By Joshua McKibben, Head of Trust, Harvey.
Harvey is investing in legal teams in public sector organizations across different countries and regions. Two recent steps are our FedRAMP Class D (High) certification in the United States and the completion of an independent IRAP assessment at the PROTECTED level in Australia. These are important steps in our work. They show our commitment to helping public sector lawyers do more for their communities and giving their organizations the security evidence they need to adopt AI with confidence.
Public sector legal work happens in many kinds of organizations. National agencies work with different departments. State and provincial teams give advice on public programs. City lawyers handle purchases and legal disputes, often with small teams. Public universities need legal support for research and education, while publicly owned utilities provide essential services. Legal teams in all these organizations help make important decisions. Their ability to use AI depends on how well it meets their work needs, responsibilities, and rules.
More Time for the Work That Supports Public Services
A public project may be delayed while a purchase agreement waits for a lawyer to review it. A request for public records may require a lawyer to read years of emails and letters. A proposed policy may require someone to compare laws with advice that is spread across different departments.
Before lawyers can give advice, they need to understand the documents, find the rules that apply, and learn what the organization already knows. This preparation takes time and attention, in addition to the legal work itself.
AI can help with the work that happens before a decision is made. For a purchase, it can compare agreements with the requirements provided, point out differences, and suggest changes for a lawyer to review. When handling a public records request, it can find documents that may be relevant and point to the original files. Lawyers remain responsible for deciding what to release, what information is exempt from release, and what is protected by legal privilege.
For investigations and court cases, AI can find facts in documents, put events in date order, compare different accounts, and point out gaps that need more investigation. This gives lawyers a clearer starting point and more time to examine the evidence, check whether an interpretation is correct, and prepare advice for the people who must make decisions.
Make Shared Legal Knowledge Useful
A legal team's ability to give advice also depends on how easily it can find and use work that has already been done. An earlier legal opinion may explain why a policy was created in its current form. An older contract may show how the parties agreed on an unusual term. Research from one department may help answer a similar question in another department.
Making earlier work easy to search, with links to the documents that support it, can give lawyers a stronger starting point when preparing new advice. They can review the earlier reasoning, check whether the law or situation has changed, and decide which parts apply to the current question.
This can help a team give more consistent advice and keep its knowledge available over time. It can also help colleagues learn from one another when people's roles and responsibilities change. The benefit comes from making past experience available for people to check and use again, while making sure access rules protect sensitive information.
Trust Must Meet Local Rules
These opportunities exist in many countries, but the rules for using AI are not the same everywhere. An Australian government department and a US government agency may have similar legal work, but they follow different rules for security, handling data, and buying technology.
Public sector organizations need clear evidence about how their information is protected, where it is handled, who can access it, and whether the service meets their responsibilities. This evidence must be useful to the organization making the decision.
Australia's Information Security Registered Assessors Program, or IRAP, is one example. An independent assessment checks the relevant security controls and gives an organization evidence it can use when assessing risk and deciding whether to approve a system. In the United States, FedRAMP covers cloud security requirements for federal government services. TX-RAMP covers requirements for cloud services used by Texas state agencies and public colleges and universities.
Security checks must also connect to the everyday work of legal teams. Lawyers need sources they can check. Administrators need ways to control who can access and share information. Teams need clear rules about review and responsibility. Together, these practices help organizations use AI responsibly in their legal work.
A Long-Term Commitment in Different Markets
Harvey's IRAP assessment checked the relevant controls in the Australian Government Information Security Manual at the PROTECTED level. It gives public sector organizations independent evidence they can use when assessing Harvey. We have also been approved as a supplier under the NSW Government ICT Services Scheme. This gives eligible buyers an established way to purchase our services.
These steps add to our FedRAMP Class D certification, completed with Knox Systems, and our TX-RAMP Level 2 certification in the United States. They show that we are investing in the different requirements of different markets and organizations.
As Head of Trust, I see this work as an ongoing responsibility. Completing an assessment gives us a starting point for discussions with an organization's security team. I expect these discussions to continue as the organization's needs and our technology change. Teams that trust us with their information should be able to understand how we maintain our security controls and respond to new risks over time. This means building security into our products and daily operations, providing clear documents, and taking responsibility for the answers we give.
Public sector legal teams should be able to expect a long-term partner that understands their responsibilities and keeps working to earn their trust. This is the commitment we are making in the markets we serve.
Learn more about Harvey for Public Sector, or visit our Trust Center to read our security documents.

–
[ This is a sponsored thought leadership article by Harvey for Artificial Lawyer. ]
