Insurance has always depended on timing. Risk is captured at a specific moment, translated into a price, and expected to remain broadly stable until the next review. For much of the industry’s history, that assumption has held. Risk moved, but not so quickly that it escaped the frame of underwriting.
That balance is now shifting. Across multiple lines, the conditions shaping exposure are becoming more fluid and less predictable. Supply chains adjust in response to disruption. Assets operate in changing environments. Digital systems evolve constantly.
The gap between when risk is assessed and what it looks like even a short time later is widening.
Technology that can provide visibility to risk in real-time is emerging as a response to that gap. Drawing on continuous data from telematics, sensors, satellite monitoring and live digital scanning, it allows insurers to move closer to the present state of risk. It does not remove uncertainty, but it reduces the distance between exposure and understanding.
“Instead of taking a snapshot of risk once a year, insurers can now see how that risk is evolving day to day,” says Peter Bammodu, Insurance Director at Earnix. “That shift is especially important in more volatile or exposed classes like aviation, marine, energy, and trade, where risk can change quickly due to geopolitical disruption, supply chain issues, or shifting operating conditions.”
What follows is a change in how underwriting is approached. “In all of these cases, underwriting is becoming less about predicting risk upfront, and more about continuously assessing it as conditions change,” Bammodu says.
The data behind that shift comes from multiple sources working together.
“It’s really the combination of data that makes the difference,” he says. “Telematics and sensors provide insight into how assets are being used and what condition they’re in. Satellite and geospatial data add context, like location, weather exposure, or proximity to risk.”
Each stream offers only a partial view. Together, they create something closer to a live picture. In commercial motor insurance, telematics programmes have reduced claims frequency by 20% to 30%, showing how continuous visibility can influence behaviour and outcomes before losses occur.
At the same time, access to data is not the same as using it effectively. “Most insurers already have access to more data than they can fully use,” Bammodu says.
“The challenge is getting that data into live decision-making. What’s changing is a move toward more connected systems, where data, models, and underwriting rules are linked.”
This changes how decisions are made. “Underwriting is becoming more responsive and more consistent, because decisions are based on what’s happening now, not just what happened in the past,” he says.
The risk beneath the record
Traditional underwriting depends heavily on what is declared. Proposal forms and historical data create a structured view of risk at a given point in time. The limitation is not that this view is wrong, but that it does not stay current for long.
“The data used to underwrite a risk and the risk itself are often not the same thing, and that gap is where losses happen,” says Melanie Hayes, Co-Founder at KYND.
This gap is especially visible in cyber insurance, where exposure changes quickly and often without clear visibility.
“A risk is assessed at a moment in time and, broadly speaking, that assessment holds until the next defined decision point,” Hayes says. “What happens in between is, in many cases, invisible. Services change, configurations drift, third-party dependencies evolve. A risk that looked one way at inception may look materially different six months in.”
Real-time intelligence brings that hidden period into view. “A feed that reflects the state of a risk at the point a decision is being made is categorically different from one built on a library that refreshes monthly,” she says. “The gap between what was true when a form was submitted and what is true now is real, and it has consequences.”
Those consequences are reflected in loss data. A large share of cyber incidents is linked to known vulnerabilities that were not fixed. Exposure often exists before an event takes place.
More data, however, does not automatically lead to better decisions.
“Most real-time data feeds tell you everything,” Hayes says. “The useful ones tell you what drives losses. The question is not what vulnerabilities exist for a given organisation, but which ones are materially exploitable and loss-correlated. Those are not the same.”
This introduces a shift in focus. The emphasis moves from collecting information to identifying what matters.
In cyber insurance, that often means prioritising signals such as exposed services, misconfigurations and actively exploited vulnerabilities. These are closer to the factors that lead to loss.
At the same time, underwriting is becoming more connected.
“Underwriters are starting to see not only the individual risk, but how it connects to others through shared technologies, suppliers or vulnerabilities, and what that means for the portfolio,” Hayes says. “Selecting a single risk also means adding to a broader exposure.”
This reflects how modern risk is structured. Dependencies are shared, and failures can spread across many organisations.
“The increasing reliance on common cloud providers, widely used software, and now AI-driven tools means that a single point of failure can affect hundreds or thousands of insureds at once,” she says.
Real-time intelligence helps reveal these concentrations earlier, before they turn into wider events.
Before the loss
The claims process has traditionally focused on what happens after an event. Real-time intelligence shifts attention to what happens before it.
“Claims is often where the conversation starts, but the biggest impact is actually before a claim happens,” says Bammodu. “Real-time monitoring allows insurers to spot issues earlier and intervene where possible.”
That intervention can take different forms. In logistics and marine insurance, it may involve identifying route changes or delays. In commercial motor insurance, it can mean recognising patterns of risky driving. In property insurance, it may involve tracking environmental conditions such as temperature or water exposure.
The principle is the same. Earlier visibility creates the chance to act before a loss occurs.
In connected insurance programmes, this approach has led to loss cost reductions of up to 25% in some portfolios. The improvement comes mainly from preventing incidents rather than handling them more efficiently.
Bammodu frames it clearly. “So while it can help speed up claims and improve fraud detection, its real value is in reducing losses in the first place and improving overall risk quality,” he says.
In cyber insurance, the same pattern applies. “The most significant impact is on loss prevention,” Hayes says. “The vulnerabilities that drive losses are frequently not novel or sophisticated. They are known, observable weaknesses that existed before an incident and remained unaddressed.”
Real-time monitoring changes when those weaknesses are identified. “When that exposure is visible before an event rather than reconstructed afterwards, the window for a different outcome exists,” she continues.
Where losses do occur, continuous data changes how claims are handled.
“For claims that do occur, continuous monitoring provides a timestamped record of conditions prior to the loss event,” Hayes says. “That changes what is contestable, compresses resolution timelines, and makes it significantly harder to misrepresent pre-existing exposure.”
These developments are also beginning to influence pricing and product design.
“Real-time data allows insurers to move toward pricing that reflects how risk actually behaves over time,” Bammodu says. “That could mean usage-based models in commercial auto, more dynamic exposure-based pricing in marine and aviation, or more responsive products in property and agriculture.”
Hayes points to the same direction while stressing the limits.
“The logical destination is pricing and policy structures that reflect current conditions rather than historical ones,” she says. “But none of that is achievable without data that is accurate enough to be trusted and complete enough to cover the full range of risks being underwritten.”
The direction of travel is clear. Insurance is moving closer to the conditions it seeks to measure, working with risk as it changes rather than relying on what was once recorded.
