Jack Williams, of Hexagon Safety & Infrastructure, made a compelling argument for the integration of Assistive AI into the ECC in this informative session. Williams started off acknowledging that many people are wary or apprehensive about incorporating AI in the communications center. Some are worried it might take over their jobs, others have perceptions shaped by such influences as the Terminator movies or other mainstream media interpretations of what AI is. Williams pointed out that AI is already in widespread use in areas such as virtual assistants (Siri, Alexa, etc.) and in search engines, as well as recommendation agents (Netflix, Spotify, etc.).
AI is defined as intelligence demonstrated by machines. There are many different types of AI, and without going too far into technical detail, Williams focused on Assistive AI as a quite useful tool in the ECC toolbox. Some of these are already being used in public safety, such as threat detection tools, image identification, video/audio transcriptions, and some cybersecurity applications. Assistive AI “assist personnel but leaves the decision-making up to humans.” In other words, it can function as an “operational blind-spot detector” or a second set of eyes for the emergency communications specialist.
The thoughtful, timely and ethical integration of assistive AI can be a force multiplier that results in better real-time decision-making. It can help capture institutional knowledge because it “learns” from every call. It can amplify intuition and help onboard new hires, and create a shared awareness with neighboring jurisdictions and other organizations. Finally, it can help rethink public safety workflows and truly assist the 9-1-1 professional, not replace them.
Williams acknowledges that the implementation of an AI tool has its challenges and passed along key principles from Rick Zack at Microsoft: finding an AI solution that operates reliably; one that treats all people the same (free from bias); a solution that protects the privacy and security of the data; incorporating a solution that connects and intentionally includes the human element; a solution that is explainable, understandable, and transparent; and finally, adopters should understand that AI does not take the place of accountability of the person using it.
Williams spent some time talking about ethical usage of AI tools and particularly the accountability aspect. Recognizing that there has yet to be case law affecting the use of AI tools; nevertheless, he is an advocate of creating strong policy and procedure to support the use of AI in the emergency communications center.
Williams encourages those who wish to know more to visit the Hexagon booth #937.
By Jennifer Kirkland
