Putting Numbers on Cyber Insurance and Cybersecurity Costs
Key Takeaways
- •Munich Re estimates the global cyber insurance market was worth nearly $15 billion last year and could reach roughly $28 billion by 2030.
- •Aon forecasts that nearly 20% of cyberattacks will involve generative AI by 2027.
- •The insurance industry's projections reflect carriers' own assessments of loss exposure, as shown by the ransomware-driven premium increases and tighter underwriting requirements of 2020–2021.
- •Current U.S. cybersecurity costs are estimated at $100 to $300 billion and could double within five years, though much of the true cost goes unmeasured.
- •The author argues that concrete market-tied numbers are essential for the AI risk debate, and the cited figures, while imperfect, are better than none.

Recent estimates put concrete figures on the growth of cyber risk. According to Munich Re's latest report, the global cyber insurance market was worth nearly $15 billion last year and is expected to reach roughly $28 billion by 2030. Aon said earlier this year that, according to its forecasts, nearly 20% of cyberattacks will involve generative AI by 2027.
These figures come from a market context — based on sectoral estimates rather than prices per se — but they offer a useful anchor (via Marc Pfeuffer). Correction: via Marc Pfeiffer. If cyber insurance expenditures are set to almost double by 2030, one might infer that cyber costs more generally could be roughly doubling as well. The projection that 20% of cyberattacks will involve generative AI does not by itself pin down losses, since that 20% could prove especially costly. Still, matching the 20% figure against the estimated near-doubling of the cyber insurance market — on the reasoning that the biggest potential losers are the most likely to buy insurance — yields sums that are very high, though arguably not the end of life as we know it.
The insurance market is itself a meaningful signal here: insurers price policies against actuarial data, and the industry has already been through one hard market cycle, when ransomware losses around 2020–2021 drove premiums sharply upward before carriers tightened underwriting requirements around multi-factor authentication, backups, and incident-response planning. Premium growth projections of this size therefore reflect carriers' own assessments of where loss exposure is heading, not just marketing optimism.
For broader context, Claude 5.1 estimates current U.S. cybersecurity costs in the range of $100 to $300 billion, a figure slated to rise considerably — potentially doubling within five years, consistent with the insurance-market trend above. The wide range of that estimate reflects a genuine measurement problem: much of the true cost of cyber incidents — downtime, reputational damage, and undetected intrusions — never shows up in any single tally.
One way to read those numbers is that cyber costs will rise about thirteen percent a year, and that defenses will eventually catch up. A more pessimistic reading is that cyber costs may keep climbing thirteen percent annually until the whole economy buckles, or until we retreat to the pre-digital era. Both optimistic and pessimistic interpretations of the figures are possible.
Whatever one's stance on AI risk debates, what is most needed is numbers — ideally tied to market data. The figures cited here may not be the right numbers, but they are better than no numbers at all. The open question remains: what other concrete numbers exist, and if none do, why not?
This post originally appeared on Marginal Revolution.
Original article: Business Insurance, via Marc Pfeiffer.