FOR LAWYERS, BY LAWYERS
Practical experiences, primers, and more for attorneys.
What is AI?
AI is a broad term that encompasses computer technologies designed to simulate human intelligence and perform tasks that typically require human-like cognition. While there is no single, universally accepted definition of AI, it generally refers to systems that can analyze data, recognize patterns, learn from experience, make informed judgments, predict future behavior, and automate functions.
At its core, AI is driven by algorithms – sets of coded instructions that enable computers to process data and execute specific tasks. These algorithms can be programmed to achieve a wide range of objectives, from simple pattern recognition to complex decision-making and content generation.
AI manifests in many different forms and applications. Some common examples include:
- Natural Language Processing (NLP): This enables computers to understand, interpret, and generate human language, as seen in legal research tools and chatbots.
- Machine Learning: This allows systems to learn and improve from experience without being explicitly programmed, as exemplified by predictive text and recommendation engines.
- Computer Vision: This enables computers to interpret and analyze visual information from the world around them, powering technologies like facial recognition and autonomous vehicles.
- Generative AI: This newer branch of AI utilizes unsupervised learning algorithms to create original content – such as text, images, audio, and code – based on patterns in existing data. Prominent examples include ChatGPT and DALL-E.
In practical terms, AI is being deployed across industries to automate and enhance a wide array of functions. Legal departments, for instance, are using AI to streamline contract review, due diligence, e-discovery, and compliance monitoring. Advertisers are harnessing AI to predict market trends and personalize recommendations. And manufacturers are integrating AI into robotic systems to optimize production and quality control.
As AI continues to evolve and permeate our personal and professional lives, it is becoming increasingly important to understand its capabilities, limitations, and implications. While AI has the potential to drive immense efficiency and innovation, it also raises complex questions around ethics, accountability, and the future of work. Navigating this landscape will require ongoing collaboration among technologists, policymakers, legal experts, and ethicists to ensure that the development and deployment of AI serves the interests of both businesses and society at large.
AI Liability
The rapid proliferation of AI in consumer products and business operations is giving rise to complex liability questions. As AI systems become more autonomous and capable of causing injury or property damage, courts and litigants are grappling with how to apply traditional tort law principles to this novel context.
In the absence of AI-specific legislation, plaintiffs have begun to test the waters with product liability claims based on theories of negligence, breach of warranty, and strict liability. These cases highlight the challenges of assigning fault when an AI-powered product is involved.
In a recent case involving GPS devices, Cruz v. Raymond Talmadge d/b/a Calvary Coach, the plaintiffs alleged that the devices were defectively designed because they directed a bus driver to follow a route under a low overpass, causing a collision. The plaintiffs claimed that the manufacturers failed to warn of this foreseeable danger and could have feasibly incorporated height restriction data. This case illustrates how traditional product defect theories may apply where the AI component is relatively limited and fault can be traced back to the original design.
However, as AI becomes more autonomous, the liability picture becomes murkier. In Nilsson v. General Motors, a motorcyclist sued the manufacturer of an autonomous vehicle (AV) that allegedly veered into his lane, causing injury. The plaintiff relied on a general negligence theory, arguing that the AV itself failed to exercise reasonable care in its driving. Intriguingly, the manufacturer admitted that the AV was required to use reasonable care, seemingly accepting the premise that the AI system could be treated as the negligent actor.
The Nilsson case, though settled before a decision on the merits, raises profound questions about the future of AI liability. Should the law treat AI as a quasi-person capable of negligence? If so, what is the appropriate standard of care – that of a reasonable human, or a new “reasonable AI” standard? How do we assess foreseeability when AI is designed to act autonomously in complex environments?
Moreover, if AI systems themselves can be liable, it remains unclear who should bear the cost of compensating victims. Some have suggested that the doctrine of res ipsa loquitur, which shifts the burden to defendants to disprove negligence, could provide a path to recovery. But this is uncharted territory for the courts.
The Legal Landscape of AI Commercial Transactions
As AI becomes increasingly integrated into business operations, products, and services, the legal landscape governing commercial transactions involving AI is evolving rapidly. Organizations looking to leverage AI capabilities must navigate a complex web of contractual issues, liability concerns, and regulatory uncertainties.
At the heart of many AI transactions are software licensing agreements, which may also involve the purchase, lease, or licensing of related equipment, services, and data. When AI is a central component of the deal, these agreements often raise unique negotiation points around risk allocation, data use, and performance guarantees.
One key area of focus is the vendor’s representations and warranties regarding the AI system’s functionality and output. Given the mission-critical nature of many AI implementations, customers will seek robust assurances about the system’s reliability and fitness for purpose. Careful drafting is needed to allocate responsibility for any failures or errors, particularly where the AI’s decision-making process is opaque.
Indemnification provisions are another crucial tool for apportioning liability in AI contracts. When an AI system’s autonomous actions cause harm, it may be unclear whether fault lies with the AI provider or the user. The parties must thoughtfully negotiate indemnity terms to ensure an appropriate balance of risk and responsibility.
Limitation of liability clauses also take on heightened importance in the AI context. The potential for catastrophic damages from AI failures, such as the shutdown of an automated production line or the breach of sensitive user data, means that liability caps must be set at levels that properly incentivize performance while providing adequate recourse for aggrieved parties.
Data rights and usage terms are another key battleground in AI transactions. AI systems rely on vast troves of data to train their algorithms and improve their performance over time. Vendors often seek broad rights to collect, aggregate, and monetize customer data across their user base. Customers, in turn, may resist such data sharing on competitiveness or privacy grounds. Finding a mutually acceptable middle ground requires careful drafting and attention to applicable data protection laws.
The rise of AI-powered consumer products, from smart home devices to self-driving cars, adds further wrinkles to the legal analysis. These products often blur the line between goods and services, raising questions about the applicability of traditional product liability doctrines. Allocation of warranty responsibilities between hardware manufacturers and software developers can also be a point of contention. And the patchwork of regulatory oversight in this space creates additional compliance challenges.
As the commercial AI market continues to mature, businesses and their legal counsel must stay attuned to the unique risks and opportunities presented by this transformative technology. Careful contract drafting, informed by a deep understanding of the technical and regulatory landscape, will be essential to unlocking the benefits of AI while mitigating potential liabilities. By proactively addressing these issues at the dealmaking stage, companies can lay the foundation for successful and sustainable AI deployments.