Public Safety Leaders Reframe AI Around Intelligence, Governance and Public Trust
Executive Q&A with I2 Group’s Roger Stokes
Artificial intelligence is reshaping public safety intelligence operations, but perhaps not in the way many anticipated. Rather than pursuing fully autonomous investigative systems, state and local agencies are increasingly focused on using AI to help intelligence professionals process larger volumes of information, strengthen collaboration across jurisdictions and accelerate operational decision-making while preserving transparency, accountability and public trust.
That evolution reflects a broader shift in how public safety organizations define intelligence itself. Success is becoming less about collecting more information and more about transforming disparate data into actionable intelligence that investigators, analysts and executive leaders can confidently use to protect communities. As AI capabilities mature, agencies are also confronting new questions surrounding governance, workforce development and the role of experienced intelligence professionals in validating AI-assisted analysis.
To better understand how these priorities are evolving, BizTechReports recently spoke with Roger Stokes of i2 Group during an executive vidcast examining the future of AI-enabled intelligence operations across state, local and education (SLED) organizations. During the discussion, Stokes explained why intelligence sharing is becoming increasingly important across public safety ecosystems, how the role of the intelligence analyst is evolving alongside AI and why governance, transparency and human judgment remain essential to building confidence in AI-assisted investigations.
The conversation also underscored a larger reality facing public safety organizations. AI may dramatically improve the speed at which agencies discover relationships, organize information and identify investigative leads, but technology alone cannot replace the experience, judgment and accountability required to transform information into defensible intelligence. Instead, AI is becoming another tool that enables intelligence professionals to work more effectively while reinforcing the governance principles that underpin public trust.
Here is what he had to say:
BTR: Artificial intelligence has become a strategic priority across virtually every industry. From your perspective, what is driving public safety organizations to make AI a strategic priority for intelligence and investigative operations?
Roger Stokes: The driving force is the growing complexity of the intelligence environment. Public safety organizations are collecting and receiving information from more systems, more agencies and more sources than ever before. At the same time, investigators and executive leaders are expected to make decisions more quickly and with greater confidence. Those two trends are occurring simultaneously, creating an environment where traditional analytical approaches alone are becoming increasingly difficult to sustain.
Artificial intelligence helps address that challenge by accelerating how information is collected, organized and correlated across multiple systems. It enables analysts to move more efficiently from raw data toward actionable intelligence by reducing the amount of time spent manually searching for relationships among people, places, events and activities. That allows investigators and operational leaders to focus more of their attention on understanding what the intelligence means rather than simply assembling it.
Equally important, AI should not be viewed as replacing the intelligence function. The purpose is not to automate investigative judgment. Rather, AI strengthens the work of intelligence professionals by providing them with faster access to relevant information while preserving their responsibility to evaluate context, validate findings and support operational decision-making. In that sense, AI becomes an enhancement to intelligence operations rather than a substitute for professional expertise.
As agencies continue integrating AI into their investigative environments, the strategic objective is not simply improving efficiency. It is improving the quality and timeliness of intelligence that ultimately supports better operational decisions across the public safety enterprise.
BTR: Beyond improving efficiency, are you seeing AI change the way agencies think about intelligence sharing across the broader public safety community?
Stokes: Absolutely. One of the most important developments is the growing recognition that intelligence extends well beyond traditional law enforcement organizations. Effective public safety depends upon information contributed by a much broader community that includes corrections, probation, parole, emergency management and other organizations that all possess valuable operational insight.
Historically, many of these organizations collected information independently and shared it only when necessary. Today, agencies increasingly recognize that fragmented information often produces fragmented understanding. AI is helping organizations overcome that challenge by bringing together information from multiple sources more quickly and presenting it within what we often describe as a common operating picture.
When investigators, intelligence analysts, executive managers and operational personnel are working from the same understanding of available information, they are better positioned to coordinate responses, identify emerging threats and make informed operational decisions. AI helps accelerate that process by integrating data that previously required substantial manual effort to collect and correlate.
That does not mean agencies abandon their individual governance responsibilities or legal authorities. Every organization maintains its own mission, statutory requirements and operational priorities. What AI enables is more effective collaboration while respecting those existing boundaries.
Perhaps the most significant philosophical shift accompanying that technology is the movement away from a traditional "need to know" mindset toward what many intelligence professionals now describe as a "responsibility to provide." If an organization possesses information that could improve public safety outcomes, there is increasing recognition that sharing that information responsibly may be just as important as protecting it.
Achieving that balance requires mature governance. Information must be shared appropriately, role-based access controls must remain intact and agencies must continue protecting privacy and civil liberties. AI can facilitate collaboration, but it must operate within governance frameworks that preserve public confidence while enabling more effective intelligence operations.
BTR: Public safety agencies are simultaneously confronting rapidly expanding data volumes and increasing expectations for faster investigative decisions. Where do you see AI having its greatest operational impact?
Stokes: I think AI is proving most valuable because it addresses both of those challenges simultaneously. Public safety organizations are dealing with unprecedented amounts of information, yet operational leaders cannot afford slower decision-making simply because more data exists. In fact, the opposite is true. They are expected to respond more quickly while maintaining the same level of analytical rigor and investigative confidence.
Artificial intelligence helps agencies organize and analyze significantly larger datasets than traditional processes could reasonably support. It enables disparate systems to contribute information more efficiently while accelerating the identification of relationships that might otherwise require extensive manual analysis.
The real operational value, however, comes from allowing intelligence professionals to spend less time gathering information and more time interpreting it.
That distinction is extremely important because AI provides speed, but it does not provide context.
Context remains the responsibility of experienced analysts who understand investigative priorities, criminal behaviors, operational environments and the relationships among multiple information sources. Those professionals determine whether identified patterns are meaningful, whether additional investigation is required and how intelligence should influence operational decisions.
In many respects, AI simply moves organizations to the point where meaningful intelligence work can begin much faster. It accelerates the mechanics of intelligence analysis without replacing the professional judgment that ultimately determines investigative outcomes.
BTR: As AI assumes greater responsibility for identifying patterns, surfacing relationships and recommending investigative pathways, how does the role of the intelligence analyst evolve?
Stokes: I actually believe the intelligence analyst becomes even more important. Technology can process enormous amounts of information, identify relationships and surface analytical observations at speeds that would be impossible through manual methods alone. What it cannot do is replace the professional judgment that experienced intelligence analysts bring to the investigative process.
Analysts remain the discipline experts. They must understand investigative methodology, operational priorities, evidentiary standards and the broader context surrounding criminal activity. It is also vital that they validate AI-assisted findings, determine whether analytical conclusions are supported by reliable information and ensure that investigative recommendations remain consistent with established governance and professional standards.
That human role becomes even more critical as agencies seek to build public confidence in AI-assisted intelligence.
Communities need to understand that AI is assisting trained professionals rather than making independent decisions that affect people's lives. Maintaining humans throughout the intelligence cycle reinforces accountability while demonstrating that experienced investigators and analysts remain responsible for the judgments that ultimately guide public safety operations.
The evolving role of the analyst also creates new professional expectations. Intelligence professionals increasingly need to understand the integrity of the data they are evaluating, the capabilities and limitations of AI technologies and the governance frameworks that regulate responsible AI use. They must be able to explain how information was obtained, how analytical conclusions were reached and why those conclusions can be trusted.
BTR: Public safety organizations often begin their AI journey by investing in technology. Beyond acquiring new analytical capabilities, where should executive leaders focus their investments to ensure long-term success?
Stokes: Technology is certainly an important part of the investment strategy, but it should never be viewed as the strategy itself. The organizations that will realize the greatest long-term value from AI are those that invest equally in the people, processes and governance that enable technology to become operationally effective.
Introducing AI into an intelligence environment should be viewed as part of a broader modernization effort. Agencies need to think about how new analytical capabilities integrate with existing investigative workflows, information-sharing processes and operational decision-making. Those considerations often determine whether technology becomes an everyday operational asset or simply another application within the enterprise.
One area that deserves particular attention is analyst career development. Intelligence analysts remain central to every stage of the intelligence cycle, yet many public safety organizations continue to experience significant workforce turnover. Recruiting talented analysts is only the beginning. Agencies also need structured programs that develop analytical expertise, expand AI competencies and create clear opportunities for long-term professional growth.
That means funding should extend beyond implementation. Organizations should invest in onboarding, continuous education and long-term sustainment so that institutional knowledge grows alongside technological capability. AI will continue to evolve rapidly, and analysts must evolve with it if agencies expect to realize sustained operational benefits.
Leadership should also recognize that intelligence modernization is an ongoing journey rather than a one-time technology deployment. New data sources, changing criminal activity, evolving governance requirements and advances in AI will continue reshaping investigative operations. Organizations that invest in adaptable people and sustainable operational processes will be much better positioned to respond to those changes than agencies that focus primarily on software acquisition.
BTR: Speed has obvious operational value, but public safety agencies must also preserve transparency, accountability and public trust. Where should leaders focus their investments to maintain that balance?
Stokes: Public trust has to remain a strategic priority throughout AI adoption. Communities need confidence that artificial intelligence is supporting responsible policing rather than replacing professional judgment or making independent decisions that affect people's lives.
That begins with clear governance. Agencies should communicate that AI is assisting intelligence professionals throughout the analytical process while experienced analysts, investigators and executive leaders remain responsible for evaluating evidence and making operational decisions. Technology accelerates portions of the intelligence cycle, but accountability continues to reside with people.
Investment should also focus on governance frameworks that help agencies operate consistently and transparently. Those frameworks need to address issues such as privacy, civil rights, civil liberties, role-based access and appropriate use of investigative data. They should establish clear expectations regarding how AI-generated information is validated, documented and incorporated into operational decision-making.
Training is another critical investment. Intelligence professionals need to understand not only how AI tools function but also their limitations. They should be capable of explaining how analytical conclusions were reached, what information supported those conclusions and where professional judgment influenced the final intelligence product.
That transparency becomes increasingly important as AI capabilities continue expanding. Organizations that openly demonstrate how human oversight, governance and accountability remain embedded throughout the intelligence process will be better positioned to maintain public confidence while responsibly introducing increasingly sophisticated analytical technologies.
BTR: Public safety operates within a highly decentralized environment where agencies maintain independent governance structures and legal authorities. How do you see AI improving collaboration across jurisdictions while still respecting those operational boundaries?
Stokes: One of AI's greatest contributions is its ability to help organizations work from a common understanding of available intelligence while respecting the governance responsibilities that remain unique to each participating agency.
Public safety has always involved multiple stakeholders. Municipal police departments, sheriffs' offices, corrections agencies, probation, parole and many other organizations all contribute information that may be relevant to an investigation. Historically, integrating those information sources has required significant manual effort, and the resulting intelligence has often remained fragmented across organizational boundaries.
AI enables agencies to bring those information sources together much more efficiently. Analysts, investigators, supervisors and executive leaders can view a common operating picture that helps everyone understand the broader investigative environment while continuing to access only the information appropriate to their specific responsibilities.
That distinction is important because successful intelligence sharing has never meant unrestricted information sharing.
Role-based access controls remain fundamental. Governance determines who may access particular information, how that information may be used and how privacy protections are maintained. AI does not eliminate those requirements. Instead, it helps agencies apply them more consistently while accelerating collaboration across multiple jurisdictions.
BTR: As public safety organizations continue integrating AI into investigative workflows, what advice would you offer leaders seeking to modernize intelligence operations while maintaining confidence in both the technology and the decisions it supports?
Stokes: My advice would be to remember that AI is a tool, not an autonomous decision-maker. There is understandable excitement surrounding artificial intelligence because it can accelerate analytical workflows, organize complex information and identify relationships much more quickly than traditional methods. Those capabilities will continue expanding, and I believe AI will become an increasingly important component of intelligence operations across public safety organizations.
However, agencies should resist viewing AI as a replacement for experienced intelligence professionals. Intelligence analysis has always depended upon judgment, context and accountability. Those responsibilities remain firmly with trained analysts, investigators and executive leadership.
Organizations should also focus on transparency. When AI contributes to an intelligence product, agencies should be comfortable explaining how technology assisted the analytical process while making it equally clear that experienced professionals evaluated the information and made the final judgments. That distinction strengthens confidence in both the technology and the intelligence it supports.
From a technology perspective, platforms such as i2 Group’s Analyst's Notebook continue evolving by incorporating AI-assisted capabilities that accelerate visualization, data integration and analytical workflows while preserving the human oversight that remains central to investigative decision-making. That evolution reflects where I believe the industry is headed. AI will continue making intelligence operations faster and more efficiently, but experienced professionals will continue determining what the intelligence means and how it should influence operational decisions.
Conclusion
Artificial intelligence is introducing significant new capabilities into public safety intelligence operations, but this interview with Roger Stokes suggests that the industry's greatest transformation may be organizational rather than technological. As agencies confront growing data volumes, increasingly complex investigations and rising expectations for interagency collaboration, AI is becoming a catalyst for rethinking how intelligence is produced, governed and shared across the public safety ecosystem.
The conversation also reinforces how organizations’s path to deriving value from AI is achieved by equipping experienced professionals with better information and more efficient analytical tools. For public safety leaders, that means the future of intelligence will depend as much on governance, workforce development and public trust as it will on advances in artificial intelligence. The agencies that successfully balance those priorities will be best positioned to deliver intelligence that is not only faster, but also transparent, accountable and worthy of the confidence placed in it by the communities they serve.
###