Dataminr sits at the intersection of the cyber and physical worlds, with a leading solution to provide companies with real-time information discovery and critical insights from more than one million public data sources. Through advanced AI and machine learning, Dataminr delivers real-time alerts that help organizations stay ahead of risks and opportunities in the physical and cybersecurity worlds.
To better understand their view on the state of the physical security market and the role of AI in mitigating risk, we sat down with Dataminr Chief Product Officer Adam Bates. Here’s what he had to say:
Using your insight from the millions of events that Dataminr ingests, what are some of the most significant physical security risks and trends facing organizations today?
From geopolitical risks to natural disasters and severe weather-related crises, it often feels like today’s world is more uncertain than ever. We recently asked Dataminr’s customers what they see as the top risks and challenges they face, and the responses ranged from long-standing concerns, including terrorism, severe weather, and supply chain disruptions, to more recent alarming trends, such as election security and threats of violence to executives. Common among all these concerns is the profound impact they can have on an organization’s people and assets — business critical in the most literal sense.
What are some of the biggest challenges organizations face in mitigating these risks?
Security leaders tell us their biggest challenge is timely awareness of the sheer volume of risk events that unexpectedly arise across their unique operational footprints across the globe, as they’re expected to have Scenario Literacy on each. However, the options for timely awareness have not traditionally been up to the challenge. Human Analysis provides rich intelligence on many facets of a risk but simply doesn’t scale to all the surface area leaders need to cover, while OSINT feeds ostensibly contain the breadth of signal required, but are notoriously difficult to operationalize.
How are you seeing these risks converge with the cybersecurity risks we also see in the landscape?
The complexity of timely awareness and Scenario Literacy is compounded when Security Leaders contemplate exposure to cyber risk — typically tens of thousands of endpoints collectively running thousands of different software applications, each exposed to a perpetually morphing set of vulnerabilities or subject to vendor updates that may inadvertently halt operations. Sometimes the impact of a disruption event is isolated to information systems, but to a growing extent, those events have physical world implications, and not just in OT/IOT scenarios. Consider last year’s Crowdstrike Falcon update as one extreme example: Security Leaders were struggling to wrap their heads around the extent of outages within their own four walls but also contending with disruption arising from unexpected outages in services their organizations depend upon — travel, shipping & logistics, employees’ ability to access routine healthcare services, etc. While an extreme example in its global reach, it exemplifies the sprawling cyber-physical implications that Security Leaders are experiencing routinely in more isolated incidents.

How can AI help organizations combat these rising risks?
The good news is that the information required to detect and understand disruption events in a timely manner (be they physical, digital, or converged) is out there in publicly available data sources. The bad news: it’s spread across 1M+ (and growing) sources generating billions of daily inputs in thousands of permutations of languages and modalities. Complexity of this scale can’t be solved with approaches that require humans to curate the output. An agentic approach is needed, and that’s exactly the route we’ve taken at Dataminr. Our real-time AI platform continuously analyzes those billions of inputs — accomplishing in one hour what it would otherwise take a team of 60 analysts working 24/7 to accomplish in a year — autonomously filtering out relevant signal from noise, detecting critical risk events across an organization’s unique risk footprint in real-time and connecting disparate signals into cohesive, continuously updating event summaries that provide actionable insight on emerging threats to the business.
What are the next advancements you expect around AI in the world of physical and cybersecurity?
Progressive organizations are actively seeking to create an environment that is instantly aware of threats and is set up operationally to respond to real-time information. As these organizations gain familiarity with the scale of agentic approaches like those employed by Dataminr, I expect they will begin to adopt them internally, automating response protocols and decision-making to a greater degree whether it be generating automated comms for detected risks in certain cases, activating war rooms with the right stakeholders and playbooks, or even deploying resources. Key to this will be a continuous flow of high-resolution, real-time risk and situational awareness information that spans the physical and cyber domains.
Insights from Adam Bates were also included in NightDragon’s recent Physical Security Market Report. See the full report here.