Cesar A. Pereira and Julian D. Ehrlich

Arguably the most impactful recent decision in the world of additional insured (AI) coverage is Burlington Ins. Co. v. NYC Tr. Auth., 29 N.Y.3d 313 (2017), issued June 6, 2017 by the New York Court of Appeals and already cited over 40 times from California to Tennessee to Kansas.

Nationwide, Burlington has prompted insurers to issue more AI tenders denials often referring to the case by name. However, recent appellate court decisions are finding these denials are frequently improper while also providing useful guidance on the new standard of AI coverage.

This discussion will examine emerging areas of certainty and uncertainty in AI coverage in the wake of Burlington with special attention to claims in the construction context.

The holding in Burlington interpreted the trigger wording found in most Insurance Services Office (ISO) AI endorsements available after 2004 which states that coverage is afforded “only with respect to bodily injury or property damage caused in whole or in part by the acts or omissions [emphasis added]” of the named insured (hereinafter “the 07 04 wording”). The same trigger wording appears in the current 04 13 forms. The court found that where the putative AI, the party seeking AI coverage, is 100 percent negligent, there is no coverage.

The court relied on the 2004 ISO commentary expressing an intent to eliminate coverage where the putative AI is solely negligent as the impetus for changing the older broader ISO “arising out of work” (hereinafter “10 01 wording”) trigger wording.

Pursuant to Burlington, to trigger coverage under the 07 04 wording, there must be more of a nexus between the named insured and the loss than was previously required under the 10 01 wording but less of a nexus than negligence by the named insured. Under this intermediate standard, the named insured must be the proximate cause of the loss.

However, causation has vexed courts since at least the time of Palsgraf v. Long Is. R.R. Co., 248 N.Y. 339 (1928), which fittingly is cited in Burlington.

Courts have long struggled to define and distinguish proximate cause from but-for cause, cause-in-fact, concurrent cause, efficient proximate cause, intervening cause, superseding cause and remote cause often stating causation is the proper province of the jury but also often ruling on causation as a matter of law. See Julian D. Ehrlich, “Consistency on Sole Proximate Cause Falls Flat,” NYLJ (Sept. 14, 2015).

Now Burlington introduces the causation challenges historically found in the liability context into the realm of coverage.

Recent decisions highlight the fact sensitive nature of causation and the importance of the pleadings as well as knowledge beyond the pleadings in AI coverage determination.

Restoring Expectations

Several recent 1st Dept. cases have helped re-establish a measure of insureds risk transfer expectations with 07 04 AI endorsements finding AI coverage while also clarifying proximate cause.

For example, Indian Harbor Ins. Co. v. Alma Tower, 165 A.D.3d 549, (1st Dept. 2018), involved a Labor Law claim brought by an injured employee against the owner and general contractor who impleaded and tendered to the employers’ insurer. The court found that the employer’s insurer had actual knowledge that their named insured “may have proximately caused the injury.”

Similarly, in Vargas v. City of New York, 158 A.D.3d 523 (1st Dept. 2018), another Labor Law case, the court found that the pleadings, which contained allegations that all defendants including the named insured caused the loss, triggered a duty to defend the putative AI.

Also, in E.E. Cruz & Co. Inc. v. Axis Surplus Ins. Co., 165 A.D.3d 603 (1st Dept. 2018), the court confirmed AI coverage for the putative AI’s “vicarious or contributory negligence.”

In cases where the appellate courts have found no AI coverage, it appears that the upstream putative AI’s defense attorneys, adjusters and investigators failed to sufficiently make a showing through extrinsic evidence that the downstream named insured, often the employer, proximately caused the loss.

Defense attorneys may have concerns that coverage is inherently rife with conflicts but they have an ethical duty to maximize insurance coverage for their clients.

Accordingly, it may be necessary for them to develop causation through investigation and discovery against employers who are not, and may never be, named parties to support theories of, for example, improper supervision, direction and control or violations of codes and regulations.


While the proximate cause standard is being clarified by the courts, many tender denials misconstrue Burlington by improperly asserting, for example:

(1) AI coverage extends only to the extent of the named insureds’ percentage of negligence;

There is simply no support in the endorsement language, ISO commentary or Burlington for a position that AI coverage is divisible such that, for example, if the named insured is 10 percent negligent, then there is only 10 percent coverage for the putative AI. It has long been the rule that the duty to defend is indivisible and does not require a determination of liability. Kaste v. Hartford Acci. & Indem., Co., 5 A.D.2d 203 (1st Dept. 1958).

This type of erroneous denial may be a variation of the argument that there is coverage only for the putative AI’s vicarious liability, a notion that was expressly rejected in Burlington. Id. at 326.

(2) The pleadings allege, or extrinsic evidence beyond the pleadings indicates, that the putative AI was negligent;

Nothing in Burlington changes the long-standing rules of policy interpretation. The duty to defend is determined by the allegations in the pleadings and extrinsic evidence against the named insured not the putative insured. Continental Cas. Co. v. Rapid-American, 80 N.Y.2d 640 (1993).

Under Burlington, coverage is triggered where the putative AI is up to 99 percent negligent.

Burlington involved the highly unusual situation where the putative AI owner admitted sole responsibility for the loss. However, typically, putative AI’s deny responsibility and negligence is hotly contested.

The courts have long said that the insurance policy is ‘litigation coverage’ in addition to liability coverage meaning the purpose of the policy is to defend allegations true, untrue, proven and unproven. Automobile Ins. Co. of Hartford v. Cook, 7 N.Y.3 131, 137 (2006).

(3) AI coverage is conditioned on a determination that the named insured is negligent;

While there are jurisdictions where “cause by acts or omissions” has been interpreted to require a finding of negligence on the named insured for AI coverage to apply, there is no such requirement in New York before or after Burlington. Moreover, the Burlington court stated that its decision “does not compel the conclusion that the endorsement incorporates a negligence requirement.”

The current 07 04 trigger is proximate cause not negligence.

However, it is axiomatic that where the named insured is 1 percent or more negligent, then the named insured is the proximate cause of the loss. “The Additional Insured Book,” Donald S. Malecki and Jack P. Gibson, 7th Ed. IRMI, p. 177 (2013).

Open Questions

Canards notwithstanding, there are valid open questions after Burlington.

For example, can the putative AI be considered solely negligent where the putative AI shares negligence with an entity other than the named insured?

Analogous case law interpreting AI endorsements with explicitly written “sole negligence” wording may support the proposition that “sole” only applies to negligence as between the putative AI and the named insured.

For example, in City of New York v. Evanston Ins. Co., 39 A.D.3d 153 (2d Dept. 2007), the court found coverage where an AI endorsement required a determination of sole negligence by the named insured and there was determination that both the named insured and a third party but not the putative AI shared negligence.

Applying this reasoning to the 07 04 where the court in Burlington effectively inferred a sole negligence component, the putative AI could still be deemed solely negligent if it shares negligence with an entity other than the named insured.

Another question is whether a claimant employee’s comparative negligence can be imputed to a named insured employer to satisfy the proximate cause requirement?

Labor Law §240 imposes strict liability on owners and general contractors meaning the claimant employee’s comparative negligence is not a defense. However, in such cases, the direct defendants may impute the claimant’s negligence to the employer under vicarious liability principles in third party claims. Hoverson v. Herbert Constr. Co., 283 A.D. 3d 237 (1st Dept. 2001).

In theory, this vicarious liability principle could similarly apply to establish that the employer proximately caused the loss through the claimant workers’ comparative negligence.

Finally, can upstream putative AI’s make allegations in third-party pleadings sufficient to trigger the AI duty to defend on the third-party defendant named insured’s policy?

The court in Indian Harbor Ins. Co. v. Alma Tower, supra, appeared to rely, at least in part, on the third-party pleading allegations against the employer in finding a reasonable possibility that the employer caused the injury. However, it is not apparent from the decision whether there was any argument made that relying on third party pleadings was improperly self-serving.

Risk Management Strategies for General Contractors

Burlington and much of its progeny arise out of construction site injuries and presents risk and opportunities for general contractors.

General contractors are the proverbial meat of the AI sandwich. Upstream, or on one side of the sandwich, Burlington presents an opportunity for general contractors and their carriers to potentially avoid risk transfer obligations to owners if an owner is 100 percent negligent for third-party claims for bodily injury or property damage. Downstream, or on the other side of the sandwich, Burlington presents a risk in that it may be used to limit a subcontractor’s and its insurer’s risk transfer obligations to a general contractor.

The advice for addressing the double edge sword of the Burlington precedence is the same for risk managers as it is for outside counsel. Burlington is a fact driven standard. When conducting incident investigations and reporting, be wary of how the facts of an incident can work to your benefit or detriment depending on whether you are seeking or attempting to avoid risk transfer.

During an investigation of an incident and in reporting it, ask if there is a way to blame an owner’s negligence 100 percent for an incident. If so, there may be an opportunity to avoid AI coverage. On the other side of the sword, ask if there is the potential that a subcontractor could blame a general contractor’s negligence 100 percent for an incident. If so, there may be a risk that a subcontractor’s insurer may attempt to avoid providing AI coverage.

Developing facts and writing up incident reports correctly may help avoid AI coverage for owners or avoid jeopardizing AI from a subcontractor’s carrier.


As the above discussion underscores, AI coverage and the 07 04 “caused by acts or omissions” wording is a work in progress in the courts.

While real questions remain, recent cases instruct that downstream insurers which misread the precise holding in Burlington to deny tenders can expect to pay to lose DJ actions and then also pay for AI coverage.

To maximize coverage for their clients, defense counsel for putative AI’s, typically owners, general counsel and their risk managers, and general counsel, will need to develop facts in support of claims that the downstream named insured proximately caused the loss to strengthen tenders.

Because the court in Burlington viewed the 2004 ISO commentary as authoritative, hopefully the next time ISO revises its AI forms, it will use clearer wording in the policy and its accompanying commentary.

Cesar A. Pereira is general counsel at OHL North America in New York. Julian D. Ehrlich is Senior Vice President Claims for Aon Construction Services Group, Commercial Risk Solutions in New York.