10 Defense Strategies in AI Litigation

Attorney Nick Oberheiden
AI Litigation Team Lead

While artificial intelligence (AI) presents many unique opportunities for both developers and licensees, it also presents many unique risks for litigation. Ai companies that develop and use generative AI are already facing disputes that involve not only novel legal issues but also issues that are not well-suited to the interpretation and application of precedent. Oftentimes, when novel legal issues arise, the courts can apply principles developed and lessons learned in prior cases to reach predictable resolutions. But, in many cases, the legal issues involved in AI litigation are so unprecedented that the district court cannot rely on their past decisions to guide them forward.

What, then, are companies to do when faced with the need to defend themselves?

Even at this early stage, we are starting to see some trends in AI litigation. Some defenses are working, while others are not. Additionally, while many disputes involve issues with essentially no applicable precedent, others do involve issues that fall within the realm of existing jurisprudence—or at least close enough to it that existing common law principles can be applied. This often makes for exceedingly complex (and interesting) litigation, and it provides companies with a unique opportunity to shape a new body of case law that could guide future AI-related litigation for decades to come.

Defending Against AI-Related Claims in Civil and Commercial Litigation

With this landscape in mind, what do company founders and executives need to know about defending against AI-related claims in civil and commercial litigation? In AI litigation, as in all types of litigation, avoiding liability starts with identifying viable defenses in light of the relevant facts and (to the extent available) relevant law. AI-related claims can vary widely in cases initiated by consumers, competitors, customers, and other parties; and, while various general defense considerations apply (as discussed below), the specific allegations at hand matter immensely. Too often, terms like “AI litigation” are used to the exclusion of more specific and more relevant subcategories of disputes. Yet, machine learning-related disputes can involve allegations ranging as widely as:

Antitrust and unfair competition claims arising out of the use of AI-based pricing algorithms or tools;
Commercial liability or wrongful death claims arising out of the use of AI-based weather forecasting or self-driving software
Employment discrimination allegations arising out of the use of AI-based human resources applications;
Medical malpractice allegations arising out of the use of AI-based diagnostic software or robotic surgical machinery; and,
Securities fraud allegations involving AI-based investing applications or companies’ failure to disclose the risks associated with their AI platforms.

Even this is just an extremely small sampling. As you can see, each of these types of cases (among many others) would require a custom-tailored defense strategy that takes into account the specific details of the plaintiff’s (or plaintiffs’) allegations. With this in mind, the following are some examples of broadly applicable defense strategies that generative AI developers and licensees may be able to use in some cases:

Relying on Contractual Protections

Comparatively speaking, one of the least novel areas of AI-related litigation may be contract-based litigation arising out of commercial agreements for AI-based products and services. While AI is forcing commercial contract lawyers to come up with new terms and conditions to an extent, many of the underlying fundamentals of commercial contracting remain the same in the large language models sphere.

As a result, companies facing AI-related claims in litigation can defend against these claims by relying on well-established contractual protections in many cases. These include (but are by no means limited to):

Limited representations and warranties
Liability waivers
Damages caps
Indemnification and insurance provisions
Integration clauses

Of course, the specific contractual protections that are available in any particular case will depend on the specific contract at issue. But, when facing an AI-related dispute that involves a contractual relationship, carefully examining the contract is one of the first steps toward formulating a strategic defense.

Reasonable Efforts to Ensure Accuracy or Protect Privacy

In some cases, companies may also be able to defend against negligence-based claims by demonstrating that they undertook reasonable efforts to ensure the accuracy of AI-generated output or outcomes. Likewise, in privacy-related litigation, demonstrating reasonable efforts to protect sensitive information can potentially serve as a defense as well. The simple fact that generative AI models produce a negative outcome does not necessarily mean that liability is warranted. A specific cause of action is necessary—and, without one, plaintiffs do not have grounds to seek legal or equitable remedies.

Here, too, contractual protections are likely to play a key role in litigation between AI developers, licensees, and their customers, employees, patients, or end users. Since AI is fundamentally a software tool, essentially all AI-related disputes that plaintiffs claim should involve an underlying license agreement. Otherwise, they may be liable for a direct copyright infringement.

Assumption of Risk

The “assumption of risk” doctrine may come into play in some AI-related cases as well. Once a company or end user accepts the terms of licensure or sale, the company or end user will be deemed to have assumed the risk of harm (whether economic or physical) in some cases. When the assumption of risk doctrine applies, it can serve as a complete defense to liability in both civil and commercial disputes.

Improper Use of the AI Tool or Platform

Likewise, if a company or end user improperly uses an AI tool or platform, then this misuse rather than the product itself may be the cause of the company’s or end user’s losses or vicarious copyright infringement. In this scenario, no liability is warranted. An analogy would be a driver seeking damages from a car manufacturer after speeding into a wall. In this scenario, the driver decides to speed, not the vehicle’s capabilities, that caused the driver’s harm. Similar defenses can arise in the generative AI litigation context in various scenarios.

Insufficient Evidence of a “Defect”

In addition to contract-based claims, some plaintiffs filed litigation against AI developers. They also assert product liability claims based on alleged defects. A “defect” has a specific definition under the law (which in some cases varies between jurisdictions). In any case, the simple fact that a product fails or causes harm does not necessarily mean that it is “defective” within the meaning of the law. If an AI platform is not defective, then a claim based on product liability is not warranted.

Asserting a Lack of Basis for Liability

Along with insufficient evidence of negligence or a product defect, a general lack of basis for liability can serve as a defense in various other circumstances as well. Given the novel nature of AI, we are seeing lawsuits alleging novel claims that seek to establish new grounds for holding AI developers and other companies accountable. If these claims lack a basis in the law, which appears to be the case in many instances, then here too, no liability is warranted.

Failure to Mitigate Damages

When AI developers and other entities cannot avoid liability entirely, they may be able to limit their liability by exposing a plaintiff’s failure to mitigate. When facing losses, parties generally have a duty to take reasonable measures to mitigate their losses to the extent possible. Once a party fails to mitigate its losses, this can effectively establish a cap on the party’s claim for damages in appropriate cases.

Dr. Nick Oberheiden

Founder

Attorney-at-Law

Lynette S. Byrd

Former DOJ Trial Attorney

Partner

Brian J. Kuester

Former U.S. Attorney

Hon. Kevin McCarthy

55th Speaker, U.S. House of Representatives (ret.)

Government Consultant

Mike Pompeo

Of Counsel

Former U.S. Secretary of State

John W. Sellers

Former Senior DOJ Trial Attorney

Linda Julin McNamara

Federal Appeals Attorney

Nicholas B. Johnson

Former Prosecutor

Roger Bach

Former Special Agent (DOJ)

Chris J. Quick

Former Special Agent (FBI & IRS-CI)

Michael S. Koslow

Former Supervisory Special Agent (DOD-OIG)

Ray Yuen

Former Supervisory Special Agent (FBI)

Defending Against AI-Related Claims in Governmental Enforcement Litigation

Along with civil and commercial litigation, AI developers and other entities may also find themselves facing government enforcement litigation related to the promotion of AI platforms or these platforms themselves. When facing civil or criminal prosecution by the U.S. Department of Justice (DOJ) or its state counterparts, companies may be able to advance defense strategies including:

Good-Faith Efforts to Comply (or Reliance on the Advice of Counsel)

Given the novel nature of AI, demonstrating good-faith efforts to comply with existing law (or reliance on the advice of counsel) can often serve as a strong defense.

Insufficient Basis for Civil Liability or Criminal Culpability

Similar to private civil or commercial litigation, if governmental enforcement litigation lacks a valid basis for civil liability or criminal culpability, then prosecution is unwarranted.

Working with the Government to Chart a Path Forward

When there are legitimate questions as to what is (or should be) permissible under the law, working with the government to chart a path forward may be an advantageous approach—both for the present and for the future.

Once again, we cannot emphasize enough that these are just 10 examples of numerous potential defense strategies in AI-related litigation involving consumers, commercial customers, governmental agencies, and other entities. In some cases, the specific facts involved will lend themselves to a particular defense. In others, the issues involved may be so novel that an entirely new line of defense is required. The key is to take an informed, forward-thinking approach in light of any relevant contractual, statutory, or common law principles, and to leverage the insights that are available to build an effective defense.

Speak with a Senior AI Litigation Attorney at Oberheiden P.C.

If your company is facing AI-related litigation, or if you have concerns about facing AI-related litigation, we invite you to get in touch. Please call 888-680-1745 or contact us online to speak with a senior AI litigation attorney at Oberheiden P.C.