ZestyAI, a risk decision platform serving the property and casualty insurance sector, has published research showing that US insurers are starting to tighten terms and revisit pricing on personal cyber cover.
After a decade in which the protection grew into a fixture of homeowners policies, AI-enabled fraud is now testing the product.
The report, titled Cyber Moves Into the Home, examined US homeowners insurance regulatory filings submitted between 2016 and 2026, including filings for standalone personal cyber products.
ZestyAI ran the analysis through ZORRO Discover, its competitive intelligence platform. The findings trace how personal cyber protection moved from a specialist add-on to a broadly offered endorsement. It now typically addresses six core exposures: cyber attacks, cyber extortion, online fraud, data breaches, cyberbullying and identity recovery.
Recent filings point to a change of strategy. Rather than lifting rates on policies already in force, carriers are launching fresh products with steeper prices and amended terms. War and cyber-war exclusions show this most clearly.
Only 5% of personal cyber programmes carried such exclusions before 2023, but that share has climbed to 71% in 2026. This is the first broad narrowing of cover after seven years of expansion. State regulators have not responded uniformly. Texas and Virginia approved the exclusions, while New York and North Carolina turned them down, so protection now varies from state to state.
The threat backdrop is intensifying. FBI figures cited in the study put reported US cybercrime losses at $20.9bn in 2025, a 26% rise on the previous year. Federal Trade Commission data shows imposter fraud losses hit $3.5bn, close to triple the 2020 figure, as voice cloning and other AI tools make impersonation schemes more convincing.
Pricing has stayed almost flat over the same period. Annual premiums for personal cyber protection typically sit near $55, and entry-level charges usually fall between $30 and $75. That band has held since 2018, even as insurers widened the endorsement and lifted maximum limits from $50,000 at launch to as high as $2m.
The report also finds that personal cyber cover became standardised more quickly than any other modern personal-lines product. Most programmes rely on common policy forms supported by reinsurers, not wording drafted by individual carriers. This let insurers bring cover to market rapidly. It also means amendments to wording and exclusions tend to sweep across the industry at once.
Carriers are also writing more precise terms for social-engineering fraud, where criminals pose as trusted people or organisations to trick victims into sending money. Such losses usually fall under existing online-fraud provisions, but newer policies increasingly define and price them as a distinct category.
ZestyAI founder and CEO Attila Toth said, “Personal cyber went from a niche endorsement to a standard part of the homeowners policy in under a decade, at a price that has barely moved since 2018. What is happening now is AI is changing the risk itself faster than any product cycle this industry has experienced before.
“We are already seeing carriers define social engineering as its own coverage, raise limits, and narrow terms around systemic events. The regulatory filings provide an early indication of how the industry is responding to these emerging threats.”
Copyright © 2026 RegTech Analyst
Copyright © 2026 RegTech Analyst





