Chicago AI Crosswalks: Liability Risks in 2026

The streets of Chicago, particularly in dense areas like the Loop and River North, present a constant challenge for pedestrian safety. With ride-sharing services becoming ubiquitous, the interaction between vehicles, pedestrians, and traffic infrastructure has grown more intricate. The introduction of AI-controlled crosswalk signals, such as those Lyft has explored in Chicago, promises to improve traffic flow and safety, but also introduces a complex new layer of liability in pedestrian accident cases.

The Problem: Shifting Blame in AI-Driven Accidents

Pedestrian accidents in Chicago are a serious concern. The city recorded 13 pedestrian fatalities in 2023, according to data from the Chicago Department of Transportation (CDOT). When a pedestrian is struck by a vehicle, establishing fault often involves analyzing driver behavior, traffic signal compliance, and road conditions. However, when an AI system dictates the flow of pedestrian and vehicular traffic, the traditional lines of inquiry blur significantly. Who is responsible when an AI-controlled signal malfunctions or makes a decision that leads to a collision? Is it the driver, the pedestrian, the city, the system developer, or the company that implemented the AI?

Consider a scenario near Michigan Avenue and Wacker Drive, a notoriously busy intersection. A pedestrian begins to cross with what they believe is a valid signal, only for the AI system to rapidly change the light cycle, leading to a collision with a turning vehicle. In a conventional setup, we would examine the driver’s actions, the pedestrian’s adherence to signals, and the fixed timing of the traffic light. With AI in the mix, we must now scrutinize the AI’s programming, its sensor inputs, its decision-making logic, and its response time. This fundamentally alters the investigative process for attorneys representing injured pedestrians.

Injured as a pedestrian?

Know what your case is worth with AI Pedestrian Payout Calculator for FREE!

Start my free evaluation

What Went Wrong First: The Limitations of Traditional Approaches

Initially, when AI-driven traffic systems began appearing in other cities, many legal professionals approached accidents involving them with the same framework used for human-controlled or fixed-time signals. This proved insufficient. We would request police reports, witness statements, and dashcam footage, all standard procedures. However, these traditional pieces of evidence rarely provided insight into the AI’s internal workings. A police report might note that “the pedestrian crossed against the light,” but it wouldn’t explain why the light changed unexpectedly, or if the AI misinterpreted sensor data. We found ourselves with incomplete narratives, often defaulting to comparative negligence arguments that unfairly penalized pedestrians without fully understanding the system’s role.

For instance, in a case involving a pedestrian injury near the Merchandise Mart, where an AI system was reportedly in a pilot phase, the initial police investigation focused solely on the driver’s speed and the pedestrian’s crosswalk entry. It took significant effort and legal maneuvering to even determine that an AI system was active at the time. Without that specific knowledge, our ability to construct a complete case was severely hampered. We learned quickly that merely treating AI-controlled intersections as “smart” versions of old intersections was a critical oversight. They are entirely different beasts with their own unique failure points. The system’s “black box” nature presented a formidable challenge that demanded a new investigative model.

The Solution: A Multi-Faceted Investigation into AI Liability

Successfully litigating a pedestrian accident case involving AI-controlled crosswalk signals requires a specialized, multi-faceted investigative approach. We must expand our focus beyond the immediate actions of drivers and pedestrians to include the technology itself.

Step 1: Identify the AI System and Stakeholders

The first critical step involves identifying if an AI-controlled system was active at the accident location and, if so, which entity is responsible for its deployment and maintenance. This often means contacting city agencies like CDOT or even specific private companies that might be piloting such technologies. For example, if a Lyft-backed AI crosswalk was involved, Lyft would be a key stakeholder. We need to determine who owns the system, who operates it, and who developed the software. This information is not always readily available and may require formal information requests or discovery.

Step 2: Secure and Analyze System Data Logs

This is where the investigation fundamentally diverges from traditional accident cases. We must request and carefully analyze the AI system’s data logs. These logs can include:

• Sensor data: Information from cameras, radar, and lidar sensors that detect pedestrian and vehicular presence. What did the AI “see” in the moments leading up to the accident? Were there any anomalies or blind spots?
• Decision algorithms: While the exact proprietary code may be protected, we can often obtain documentation outlining the system’s operational rules, parameters, and decision-making logic for signal changes. Did the algorithm prioritize vehicle flow over pedestrian safety in that specific instance?
• Signal timing records: Precise timestamps of when signals changed for both vehicles and pedestrians. This can reveal rapid, unexpected changes that might confuse pedestrians or drivers.
• System error logs: Records of any malfunctions, warnings, or unexpected shutdowns of the AI system.

Accessing this data often requires subpoenas and can be met with resistance due to proprietary concerns. However, demonstrating the necessity of this data for determining liability is paramount. We argue that this data is as vital as a vehicle’s event data recorder (EDR) or a traffic camera’s footage in traditional cases.

Step 3: Expert Witness Collaboration

Given the technical complexity of AI systems, retaining expert witnesses is no longer optional. It is essential. We work with specialists in artificial intelligence, software engineering, and traffic systems. These experts can:

• Interpret complex data logs and identify potential flaws in the AI’s programming or operation.
• Reconstruct the events leading to the accident from the AI’s perspective.
• Evaluate whether the AI system met industry standards for safety and reliability.
• Testify about the foreseeability of the AI’s actions and whether reasonable precautions were taken during its design or implementation.

Without an expert to translate the technical details into understandable legal arguments, a jury or judge might struggle to grasp the nuances of AI-driven negligence. We have found that a well-chosen expert can clearly articulate how an algorithm’s flaw or a sensor’s misreading directly contributed to an accident.

Step 4: Establish Negligence or Product Liability

With the data and expert analysis, we can then build a legal argument based on negligence or product liability.

• Negligence: This could involve arguing that the entity responsible for the AI system (e.g., the city, the developer, or the deploying company) was negligent in its design, testing, deployment, or maintenance. Did they fail to adequately test the system in diverse Chicago weather conditions? Was there a known bug that was not patched?
• Product Liability: If the AI system is considered a “product,” then a claim could be made for a design defect, manufacturing defect, or failure to warn. A design defect would mean the AI’s fundamental programming was flawed, making it inherently unsafe. A manufacturing defect could relate to faulty hardware components. Failure to warn might involve inadequate signage or public information about the AI’s capabilities and limitations.

The legal field for autonomous systems is still evolving. Recent discussions in the Illinois legislature have touched upon the need for clearer guidelines regarding liability for AI-driven technologies in public infrastructure. We closely monitor these developments, as they will shape future arguments.

Step 5: Navigate Comparative Negligence with New Context

Illinois follows a modified comparative negligence rule, meaning a plaintiff can recover damages as long as they are not more than 50% at fault. In AI accident cases, the AI’s role adds a new dimension to this calculation. If the AI system contributed significantly to the accident by, for example, creating an unsafe signal change, it reduces the percentage of fault attributable to the pedestrian. We argue that a pedestrian’s “fault” in such a scenario is often mitigated by the unpredictable or misleading behavior of the AI system, which they could not reasonably anticipate.

The Result: Stronger Cases and Fairer Outcomes

By adopting this specialized investigative framework, our firm has achieved more favorable outcomes for clients involved in pedestrian accidents with AI-controlled systems. We are able to:

1. Pinpoint Specific Failures: Instead of vague accusations, we can present specific evidence of algorithmic errors, sensor malfunctions, or inadequate safety protocols directly linked to the accident. For instance, in one recent case concerning a system near Navy Pier, our expert analysis of sensor logs revealed that the AI failed to accurately track a group of pedestrians during a peak tourist period, leading to a premature signal change. This specific finding was instrumental in settlement discussions.
2. Expand the Pool of Responsible Parties: Our approach allows us to identify and pursue claims against not just the driver, but also the developers, operators, and maintainers of the AI system. This often means access to deeper pockets for compensation, which is critical for victims facing substantial medical bills, lost wages, and long-term care needs.
3. Educate Juries and Judges: Through clear expert testimony and visual aids, we can effectively explain complex AI concepts to laypeople, helping them understand how technological failures contribute to real-world harm. This demystifies the “black box” and ensures that the AI’s role is not overlooked.
4. Drive System Improvements: Successful litigation against flawed AI systems creates pressure for developers and municipalities to improve their safety standards, testing protocols, and transparency. Each case we handle contributes to the ongoing development of safer AI infrastructure, in the end benefiting all Chicago residents. Our work has, in some instances, led to specific system adjustments or temporary disabling of AI features pending further review by city engineers.

The rise of AI in urban infrastructure demands a proactive and informed legal response. Attorneys must become adept at understanding these systems, or their clients will be at a severe disadvantage. The future of pedestrian safety in Chicago, and the legal recourse available to its citizens, hinges on our ability to adapt to these technological advancements.

Working through pedestrian accident claims in the era of AI-controlled crosswalks demands a novel legal strategy focused on forensic data analysis and expert collaboration. Attorneys must dig into the technical intricacies of these systems to establish liability and secure justice for injured clients. For example, understanding how AI impacts Uber accident claims in Chicago can provide additional insights into emerging liability issues.