Cyber insurance requirements: what regulators publish and what buyers must show, outlook to 2027
State regulators set how insurers price cyber risk and secure their own data, while Congress decides the future of the federal terrorism backstop that cyber losses only partly fit.
The question for 2027
Cyber insurance has become one of the places where security controls are tested outside an examination. Underwriters ask what an organization does about access, patching, backups and vendors before they quote, and the answers set the price, the limit and sometimes whether cover is offered at all. Regulators shape that exchange from two directions: state insurance regulators supervise how insurers measure and price cyber risk, and federal bodies are deciding what happens to losses too large for the private market. This outlook sets out what those regulators have published, as checked on November 16, 2026, and what it means for the controls banks, health systems and public agencies are asked to show.
What the NAIC's 2026 report shows
The NAIC's 2026 Report on the Cybersecurity Insurance Market covers the 2025 reporting year and draws on the Cybersecurity Insurance Coverage Supplement that insurers file with their annual statements, plus alien surplus lines data from the NAIC's International Insurers Department [1]. Total US direct written premium, including alien surplus lines, was about $9.46 billion in 2025, up about 3.5% from $9.14 billion in 2024; domestic insurers wrote $7.46 billion, against $2.75 billion in 2020 [1]. The NAIC announced the report on October 21, 2026 under the headline that the market has nearly tripled since 2020 [2].
Volume grew faster than premium. Domestic policies in force rose 33% to 5.84 million and reported claims rose 20% to 56,323 [1]. Cyber endorsements to other commercial policies made up nearly two-thirds of policies in force [2], while primary policies carried about 64% of premium and endorsements about 4% [1]. The 20 largest groups in the report's table held a cumulative 76.82% market share [1]. The report does not state a count of insurers writing cyber cover, and we have not inferred one.
Results are moving the wrong way for insurers. NAIC reports, citing AM Best, that the industry cyber loss ratio rose to 53 in 2025 from 48.7 in 2024, and that the first quarter of 2026 was the eighth consecutive quarter of US cyber rate cuts; its November 2026 briefing credits both figures to AM Best, the rate trend to broker pricing data via AM Best [1] [3]. The primary book's loss and defense and cost containment ratio rose about 8 points to 57%, and the endorsement ratio nearly doubled from 10.2% to 19.1% [1]. The NAIC's own November 5, 2026 briefing to its Cybersecurity (H) Working Group cautions that several 2026 findings had to come from outside sources, and that the supplement does not show severity by insured size, pricing trends, or reinsurance and net results [3]. Its list of five gaps raised by industry analysts includes no state-level financial data and no quarterly data [3].
Moraph's view
Moraph's view: the NAIC report calls rising losses under falling prices the inverse of 2021, when a ransomware surge brought an immediate price correction. We expect any correction this time to arrive at renewal as tougher questions before it arrives as higher rates, because an insurer can tighten its application faster than it can win a rate filing or reprice a book. Buyers with large, complex estates should expect the most scrutiny, since the report expects severity to rise most sharply for large-account losses.
The rulebook for insurers' own security
The NAIC Insurance Data Security Model Law (Model #668) sets security duties for insurers, producers and other licensees, and those duties mirror what insurers now ask of their customers [6]. It requires a written information security program based on a risk assessment, board oversight, due diligence on third-party service providers, an incident response plan and, for insurers domiciled in the state, an annual compliance certification to the commissioner by February 15 [6]. Among the security measures it lists are effective controls, which may include multi-factor authentication, for anyone accessing nonpublic information, and regular testing and monitoring to detect attacks [6]. A licensee must notify the commissioner no later than 72 hours after determining that a cybersecurity event occurred, where the state is its home state or, in set cases, the event involves 250 or more of the state's residents [6]. Licensees with fewer than ten employees are exempt from the program section, and a HIPAA-covered licensee that keeps a HIPAA security program and certifies compliance is treated as meeting it [6].
The NAIC's implementation map, status as of November 3, 2026, shows 28 jurisdictions as having adopted the model, one pending, and New York with other insurance data security provisions [4]. The April 1, 2026 version of the map showed the same 28 adopted and one pending [5]. The November map shows the pending marker on the District of Columbia and leaves several of the largest insurance markets, including California, Texas and Florida, outside the adopted group [4]. NAIC notes that the map does not decide whether a state's law contains every element of the model [4].
New York's framework for pricing cyber risk
New York took a different route. The Department of Financial Services issued Insurance Circular Letter No. 2 (2021), the Cyber Insurance Risk Framework, on February 4, 2021, addressed to all authorized property and casualty insurers [7]. It asks insurers that write cyber cover for a board-approved cyber insurance risk strategy and sets out practices that include managing silent cyber risk, evaluating systemic risk, rigorously measuring insured risk, educating insureds and producers, obtaining cybersecurity expertise, and requiring notice to law enforcement [7]. DFS also recommended against paying ransoms in the letter [7].
For buyers, the framework's measurement section is the important passage. It says assessing an insured commonly starts with surveys and interviews on corporate governance and controls, vulnerability management, access controls, encryption, endpoint monitoring, boundary defenses, incident response planning and third-party security policies, and that insurers should compare that information with past claims to identify the risk tied to specific control gaps [7]. It also tells insurers to understand the critical third parties their insureds use and to model a catastrophic event at such a provider [7]. As checked on November 16, 2026, the letter remains posted on the DFS site. DFS's September 10, 2026 industry letter on risk assessments under Part 500 says it creates no new obligations and that risk assessments must be reviewed at least annually and when a material change occurs [8].
Moraph's view
Moraph's view: the DFS list of underwriting topics has become, in practice, the outline of most cyber applications. An organization that can answer each heading with evidence (configuration exports, test results, a vendor inventory) rather than yes or no answers is in a stronger position at renewal, whatever state its insurer sits in.
The federal question: who pays for a catastrophe
The Terrorism Risk Insurance Program can reimburse insured losses from a cyber attack certified as an act of terrorism when the cover is written in an eligible line, and Treasury began collecting cyber data in its program data calls in 2018 [9]. Treasury's June 2026 effectiveness report records TRIP-eligible cyber direct earned premium of $5.7 billion in 2025, down from $6.2 billion in 2024, with non-small insurers holding 67% of that market [9]. The same report says commenters identified uncertainty over certifying cyber events as terrorism and possible large protection gaps for catastrophic incidents that are not certified [9].
GAO's June 21, 2022 report found that insurers were limiting their exposure to systemic events, for example by excluding cyber warfare and infrastructure outages, and recommended that the Federal Insurance Office and CISA jointly assess whether catastrophic cyber risk to critical infrastructure warrants a federal insurance response [12]. Treasury sought comment on a potential federal insurance response on September 29, 2022, with comments due November 14, 2022 [10]. GAO's April 2026 status note says Treasury received 55 responses, completed an initial assessment in March 2024, held an industry conference in May 2024, sought comment and data on cyber losses in March 2026, and had not told GAO when it would conclude the assessment; both recommendations remain open [12]. GAO's September 23, 2025 statement on reauthorization adds that cyberattacks may not meet TRIA's requirements that an attack be violent or dangerous to human life, property or infrastructure, be part of an effort to coerce the US population or government or influence policy, and cause damage in the United States or specific areas outside it [13]. The March 25, 2026 Federal Register request for comment that fed Treasury's report closed on May 8, 2026 [11].
TRIP expires on December 31, 2027 under the 2019 reauthorization [9]. The House passed H.R. 7128 on June 29, 2026 by 373 to 15, and the Senate received it on July 13, 2026 [15]. That bill would extend the program through 2034 and set deadlines for Treasury's certification decisions, including a 90-day review after a Federal Register notice that may be extended to no more than 365 days after the damage [16]. The Senate passed its own bill, S. 4395, by unanimous consent on September 28, 2026; it extends the program to 2034 without the certification changes, and its latest action as checked is "Held at the desk" in the House on October 5, 2026 [17] [18].
Moraph's view
Moraph's view: an extension of TRIP is likely before the deadline; a federal backstop for non-terrorist catastrophic cyber events is not. Whichever bill prevails, certification deadlines matter for cyber more than for physical attacks, because attribution takes time. For buyers, the practical result through 2028 is that systemic events stay largely outside both private cover and federal support, so resilience against a cloud, software or provider outage has to be planned as uninsured.
What this means for the controls you are asked to show
Bank regulators do not require cover. The FFIEC's 2018 joint statement, issued to national banks by the OCC on April 11, 2018, says its members do not require institutions to maintain cyber insurance and that cover does not diminish the importance of a sound control environment [19]. Banks therefore meet insurer questions as a second assessment of controls their examiners already review; differences between the two answers are a supervisory and a coverage risk at once.
Health systems meet the market where loss pressure is high. GAO's 2021 market report said insurers had reduced limits for some sectors, such as healthcare and education [14]. The NAIC report says deepfake fraud places renewed emphasis on payment verification and out-of-band authentication, and that "harvest now, decrypt later" attacks raise the importance of post-quantum cryptography readiness and data retention practices as underwriting criteria [1].
Public agencies face the systemic question most directly. The FIO and CISA assessment GAO asked for concerns catastrophic risk to critical infrastructure [12], and the NYDFS framework already tells insurers to model the failure of a critical provider their insureds share [7]. Agencies that rely on a small number of cloud and software providers should expect questions on dependency mapping and recovery from provider outages, alongside the standard access, patching and backup questions.
What this means for you
- Banks: Keep one evidence set for examiners and underwriters. Cover is not required by the FFIEC members [19], but application answers on access, patching and vendors should match what Part 500 or examination files show, since DFS expects risk assessments refreshed at least annually [8].
- Health systems: Prepare for limit and sublimit pressure: GAO reported insurers cutting limits for healthcare [14], and the NAIC report says deepfake fraud places renewed emphasis on payment verification and out-of-band authentication [1]. Document callback and approval controls for funds transfers before renewal.
- Public agencies: Map dependence on shared cloud and software providers and test recovery from their outage; insurers are told to model such events [7], and no federal backstop for non-terrorist catastrophic cyber losses exists as GAO's recommendation remains open [12].
- Insurers and producers: Track state adoption of the data security model and the 72-hour commissioner notice it sets [6], and the TRIP reauthorization bills, whose certification deadlines differ between the House and Senate versions [16] [18].
Moraph predictions
Every prediction is Moraph's view, not a fact. Each has a date, a confidence level (70% means we expect to be wrong about three times in ten) and a test anyone can apply. We score them on this page when the date comes and never remove one. How we make and score predictions.
P1 By Confidence 85% Open
A law extending the Terrorism Risk Insurance Program beyond December 31, 2027 is enacted on or before December 31, 2027.
How we will check: Congress.gov or govinfo shows a public law, enacted on or before December 31, 2027, that amends section 108(a) of the Terrorism Risk Insurance Act of 2002 to set a termination date after 2027.
Basis: The program expires December 31, 2027 [9]; the House passed H.R. 7128 by 373 to 15 [15] and the Senate passed S. 4395 by unanimous consent [17], both extending it to 2034.
P2 By Confidence 70% Open
GAO still lists its 2022 recommendation that the Federal Insurance Office assess a federal insurance response to catastrophic cyber incidents as open on December 31, 2027.
How we will check: The recommendations section of GAO-22-104256 on gao.gov shows the Treasury (Federal Insurance Office) recommendation with status Open on December 31, 2027.
Basis: As of April 2026, Treasury had not told GAO when it would conclude its assessment or report to Congress [12], and both recommendations, made in June 2022, remain open [12] [13].
P3 By Confidence 75% Open
NAIC's report covering the 2026 reporting year shows more than 6.0 million domestic cyber policies in force.
How we will check: The NAIC Report on the Cybersecurity Insurance Market for the 2026 reporting year, published on or before December 31, 2027, states domestic policies in force above 6.0 million.
Basis: Domestic policies in force rose 33% to 5.84 million in 2025, driven by endorsements, which rose 53.8% [1].
P4 By Confidence 55% Open
NAIC's Model #668 implementation map shows at least 29 jurisdictions as having adopted the Insurance Data Security Model Law.
How we will check: The NAIC implementation map for Model Act #668, with a status date on or before December 31, 2027, lists 29 or more adopted jurisdictions in its legend.
Basis: The map showed 28 adopted and one pending on April 1, 2026 [5] and the same on November 3, 2026 [4].
Facts on this page link to their sources. Text marked "Moraph's view", and every prediction, is our own analysis, not a fact or advice. Methodology.
Spotted an error? Tell us and we will check it and correct the piece where needed.