This article is by Emmanuèle Lutfalla, Partner, and Louis Fer, Associate, at Signature Litigation Paris
Insurance has always been a mirror of society’s risks. When economies were industrial, insurers covered factories and ships. When mobility expanded, motor insurance became dominant. Today, as digital technologies reshape everyday life, the insurance industry faces a new inflection point.
Artificial intelligence, autonomous systems, connected devices, and augmented human technologies are not simply generating fresh categories of exposure. They are altering the very nature of what risk means. Losses are increasingly intangible, responsibility is shared between humans and machines, and causation is harder to prove.
For insurers, this is more than a technical adjustment. It is a strategic challenge that calls into question underwriting models, policy design, and even the traditional role of insurance itself.
The industry is moving from insuring objects to insuring behaviours embedded in code.
The Rise of Invisible Exposures
In the past, risk was something one could see and measure. Buildings burned, machinery broke, vehicles crashed. The damage was physical and observable.
Digital technologies disrupt this logic. Today, a company’s most valuable assets may be data sets or algorithms. A failure may consist of corrupted information, biased outputs, or automated decisions that produce legal or reputational harm. Nothing tangible is destroyed, yet the financial consequences can be enormous.
These “invisible exposures” challenge the assumptions underlying conventional coverage.
When an AI system denies credit unfairly or disseminates harmful advice, what exactly has gone wrong? Is it a defect, an error in design, or simply the emergent behaviour of a learning system? The answer determines whether a loss is insurable and under which policy.
Cyber-insurance has rapidly expanded to address data breaches and ransomware attacks, but the next generation of digital risks goes further. It concerns not just security failures but the everyday functioning of algorithms themselves. The line between operational performance and liability becomes blurred.
For underwriters, pricing these risks is notoriously difficult. Historical data is rare, incidents are heterogeneous, and the technology evolves faster than actuarial models can adapt. Traditional probability-based methods struggle when the risk landscape changes continuously.
Prevention Becomes Part of the Product
Faced with this uncertainty, insurers are redefining their value proposition. Rather than waiting for claims, they are increasingly intervening upstream.
Coverage is often linked to preventive measures: cybersecurity audits, compliance checks, governance frameworks, and technical standards. Companies seeking insurance must demonstrate robust controls over their algorithms, data management, and system resilience.
In practice, insurance becomes intertwined with risk management consulting. This shift alters the relationship between insurer and insured. The insurer is no longer just a financial backstop but a partner in shaping operational practices. By conditioning coverage on certain safeguards, insurers effectively set market standards for digital responsibility.
The policy evolves from a passive promise of indemnification into an active tool of discipline. From a business perspective, this approach also protects insurers themselves. By improving prevention, they reduce volatility and make otherwise uninsurable risks more manageable. Prevention and profitability become aligned.
When Hardware Meets Software
While some risks are purely digital, others arise at the intersection of code and physical assets. Connected and autonomous vehicles represent the most visible example.
Cars are increasingly controlled by software: sensors interpret the environment, algorithms make driving decisions, and updates are delivered remotely. A modern vehicle resembles a mobile computer as much as a mechanical machine.
When accidents occur, the cause may lie anywhere in this complex system. Was it human distraction, faulty hardware, or flawed software logic? Each possibility points to a different responsible party.
This complexity transforms claims handling.
Instead of straightforward accident reports, insurers must analyse technical logs and software performance. Liability may be shared between drivers, manufacturers, and technology suppliers. As a result, a single incident can trigger multiple insurance policies and lengthy subrogation disputes.
Yet, interestingly, the industry has not abandoned traditional motor insurance. The physical crash remains the central event around which coverage is organised. Digital elements are layered onto existing frameworks through endorsements and exclusions.
This incremental approach allows insurers to innovate without dismantling established structures.
Learning from Past Adaptations
Insurance has encountered similar transitions before. Professional liability, environmental risk, and product recall coverage all required the industry to extend beyond purely material damage. In each case, insurers responded by refining definitions, adding riders, and developing specialised products rather than reinventing the entire system.
Today’s digital risks follow a comparable trajectory.
By anchoring policies to tangible triggers: an accident, a breach, a measurable loss, insurers maintain legal clarity while accommodating new exposures. Flexibility, rather than rupture, has been the dominant strategy.
However, this strategy may have limits.
The challenge of systemic and interconnected risk
The next wave of technology could stretch existing models further than ever before. As AI systems connect entire networks, smart cities, automated supply chains, cloud-based platforms, risks become systemic. A single software failure could simultaneously affect thousands of users or devices.
These cascading losses do not fit neatly into traditional categories. They resemble natural catastrophes more than isolated incidents, yet their origin lies in human-designed systems.
For insurers, systemic digital risk raises uncomfortable questions about accumulation and insurability. How can exposure be capped when one event may trigger claims across multiple sectors at once? How should responsibility be allocated when harm results from the interaction of many independent systems?
Such scenarios may require new legal and financial architectures, perhaps involving pooled mechanisms, public–private partnerships, or parametric solutions.
Redefining the future role of insurance
Ultimately, emerging technologies are forcing insurers to rethink their identity. The industry can no longer rely solely on compensating losses after they occur. It must understand software, data science, and system design. It must anticipate risks that are dynamic, distributed, and often invisible.
In this environment, insurance becomes less about repairing damage and more about enabling trust. By setting standards, incentivising best practices, and absorbing residual uncertainty, insurers help make technological innovation socially acceptable.
The challenge is significant, but it also represents an opportunity. Those who adapt quickly, by combining legal expertise, technical literacy, and proactive risk management, will shape the next chapter of the industry.
As society moves from accidents to algorithms, insurance law and practice must evolve accordingly. The future of insurance will not be written solely in courtrooms or contracts, but increasingly in code.
Be the first to comment