4 Ways AI Is Changing the Public Safety Threat Landscape
Artificial intelligence (AI) is reshaping the threat landscape for public safety organizations. While threats such as fraud, trafficking, extremist activity, influence operations, and physical threat planning are not new, threat actors around the global now have access to generative AI (GenAI) tools that enable them to conduct these activites faster, at greater scale, and with less effort.
Rather than creating entirely new threat categories, AI is accelerating and amplifying many existing threat activities. Recent research conducted by the Center for Internet Security (CIS) points to four key ways AI is changing the public safety risk environment.
1. AI Can Lower the Barrier to Action
Activities that once required significant time, expertise, or research can now be supported by widely available AI tools.
CIS testing of publicly available AI platforms found that carefully crafted prompts could produce information relevant to physical threat planning despite existing safeguards. While much of this information is available through traditional research methods, AI makes it easier to find, organize, and synthesize.
Subsequent testing that specifically examined AI-enabled threats to event and stadium security produced similar findings. Each tested model in this study was able to be prompted to generate information about targeting emergency services.
The same dynamic applies across other threat areas. Across a wide array of criminal operations, including terrorism, narcotics trafficking, human trafficking, and others, emerging reporting reviewed by CIS suggests AI can support illegal activities through planning, surveillance, impersonation, financial transactions, and operational efficiency.
For public safety organizations, the concern is straightforward: AI can reduce the time, effort, and, in some cases, expertise required to support harmful activity.
2. AI Increases Speed and Scale
GenAI enables users to create, analyze, and distribute information far more efficiently than manual processes allow. That means generating dozens of versions of a message, translating content into multiple languages, or analyzing large amounts of publicly available information can happen faster and at greater scale.
These capabilities have implications across the threat landscape. CIS research on human trafficking identified potential uses of AI for personalized communications, translation, impersonation, and other deceptive outreach tactics. Similarly, CIS research on drug trafficking analyzed how AI is and can potentially be utilized to augment existing digital tools and operational activities for planning, communications, financial transactions, surveillance, and efforts to evade law enforcement. Even where adoption remains limited, AI can expand the reach and efficiency of existing illicit tactics.
At the same time, AI-generated content creates challenges for analysts and investigators. As AI makes content easier to generate, the volume of information, leads, and claims that must be evaluated by public safety organizations increases as well.
3. AI Makes Deception More Convincing
GenAI has also improved the ability to create realistic text, images, audio, and video.
For public safety, the concern extends beyond whether a piece of content is real or fake. What often matters most is how people respond to it.
CIS research on synthetic media and large-scale public gatherings examined scenarios involving fabricated emergency messages, false security incident reports, and impersonation of public officials, event organizers, broadcasters, and other trusted sources. Threats may involve not only fully synthetic deepfakes but also hybrid manipulations that alter authentic content in ways that may be harder to detect and easier to weaponize.
A convincing false report of an active threat or evacuation order could influence crowd behavior, overwhelm emergency services, or complicate response efforts. Even short-lived deception can create real-world consequences. For example, AI-generated audio, video, or text may be used to report fabricated threats, such as an active shooter, prompting large-scale law enforcement responses.
This makes trusted communications, established information sources, and rapid verification of information during incidents increasingly important.
4. AI Blurs Traditional Threat Boundaries
AI-driven threats rarely fit within a single public safety discipline.
A synthetic video distributed online can become a physical security issue. AI-assisted research can support physical threat planning. Personalized communications can facilitate fraud or victim targeting. Threat actors can rapidly synthesize information collected from public sources to support criminal activity.
Major events often make the convergence of these threat environments especially visible. During the preparations for and throughout the FIFA World Cup 2026, CIS monitored threats spanning physical, cyber, and information environments. Analysts observed threats of violence, swatting attempts, extremist messaging, online scams, foreign information operations, and drone activity. AI adds another layer of complexity because the same capabilities can support activity across several of these threat areas.
In today’s multidimensional threat environment, threats may quickly cross traditional organizational boundaries. As a result, intelligence, law enforcement, cybersecurity, emergency management, and event security teams may all need to respond to different aspects of the same incident.
What Public Safety Organizations Should Take From This
Public safety organizations should incorporate AI considerations into threat assessments, planning efforts, exercises, and investigations.
The key question is not whether AI creates entirely new threats, but how it changes existing ones. Can it make research faster, impersonation more convincing, fraud more scalable, or information operations harder to counter?
It also increases the importance of information sharing. As AI alters the traditional movement of information across physical, cyber, and online environments, organizations must be able to connect indicators and intelligence from multiple sources. Different organizations may be better equipped to identify different threat indicators, such as a suspicious online narrative, an account connected to previous threat activity, or activity at a physical location. Connecting those pieces becomes more important as AI increases the speed at which information can move across the threat environment.
Public safety organizations will not be able to predict every future use of AI. A more practical approach is to integrate AI considerations into existing security practices, while focusing on how AI can make threats faster, broader, more automated, and more deceptive. This is consistent with broader CIS guidance, which emphasizes applying established security fundamentals to AI while adapting them to the technology’s unique capabilities and risks.
CIS has developed resources to help organizations take that approach, including An Introduction to Artificial Intelligence, which examines AI through the lens of established security fundamentals, and the Artificial Intelligence (AI) Agents Companion Guide, which provides practical guidance for applying the CIS Critical Security Controls to AI agents and their associated risks.
AI can also strengthen defensive capabilities. Through the AI Cyber Defense Pilot, CIS, OpenAI, and the Multi-State Information Sharing and Analysis Center (MS-ISAC) are examining how state and local governments and critical infrastructure organizations can use AI to identify risks, prioritize security actions, improve cyber hygiene, and strengthen readiness.
The challenge for public safety organizations is therefore twofold: understanding how AI can amplify existing threats while also identifying where it can strengthen prevention, detection, investigation, and response.
As of June 23, 2025, the MS-ISAC has introduced a fee-based membership. Any potential reference to no-cost MS-ISAC services no longer applies.