NAPD Virtual Tech Expo 2026

Artificial intelligence is already reshaping criminal defense practice—not as a future hypothetical, but as a present force driving both the expansion of law enforcement surveillance and the overwhelming growth of defense discovery. This two-part training will help defenders understand how AI-enabled tools are changing the evidentiary landscape and how defense teams can respond thoughtfully, ethically, and effectively.  
Part 1: Law Enforcement AI and the Discovery Explosion
The first session will examine how law enforcement agencies are using AI-enabled tools to collect, analyze, organize, and act on increasingly large volumes of data. Using Axon as a case study, we will explore how body-worn camera ecosystems, automated transcription, video analytics, report-writing tools, and related platforms are changing the scale and character of criminal discovery. The session will also address the constitutional, statutory, and practical implications of these technologies, including questions related to surveillance, search and seizure, confrontation, due process, disclosure obligations, reliability, and defense access.

Artificial intelligence is already reshaping criminal defense practice—not as a future hypothetical, but as a present force driving both the expansion of law enforcement surveillance and the overwhelming growth of defense discovery. This two-part training will help defenders understand how AI-enabled tools are changing the evidentiary landscape and how defense teams can respond thoughtfully, ethically, and effectively.  
Part 2: Defense AI and a Roadmap for Responsible Use
The second session will turn to the defense response. As AI-enabled policing produces more data than defense teams can realistically review using traditional methods, defenders must develop principled ways to evaluate whether, when, and how AI tools can assist with casework. This session will provide a roadmap for thinking through those decisions, beginning with the actual defense problem to be solved rather than the tool itself. We will consider potential defense uses of AI for discovery review, issue spotting, organization, summarization, and litigation support, while also addressing confidentiality, accuracy, bias, privilege, protective orders, client consent, supervision, and ethical obligations. The goal is not to promote AI adoption for its own sake, but to equip defenders with a practical framework for making informed, strategic, and rights-protective decisions in an AI-saturated criminal legal system.

AI is changing how lawyers research, write, and analyze the law. It can save busy practitioners time. It can also help lawyers who are not strong writers and researchers create better documents. But AI's quality of research and analysis is often unable to perform well on nuanced, complex, or emerging issues, and it is especially weak at analyzing matters of first impression. In this session, I will take an issue involving confrontation rights at a state community supervision revocation hearing, and show you the strengths and weaknesses of AI, using multiple platforms. In the end, I argue that what is best for clients and the development of the law may be a blend of AI and traditional legal research and analysis, with emphasis on the latter.

AI tools now touch nearly every stage of a criminal case, from investigation through sentencing. These tools are marketed as neutral, but the data and design choices behind them are not. This session breaks down where algorithmic bias enters these tools and how it moves from code to consequence, covering facial recognition, AI-generated incident summaries, and pretrial risk-assessment instruments. Attendees leave with a working framework for recognizing algorithmic bias, so justice, not the algorithm, has the final word.

Prosecutors regularly rely on AI e-discovery software, known as technology assisted review (TAR) tools, to sort and prioritize digital evidence. These tools implicate constitutional concerns: they can either risk suppressing or help to surface exculpatory and impeachment evidence that prosecutors must disclose under the Brady due process rule. Yet doctrine, agency guidance, and scholarship offer virtually no direction on this problem. This presentation examines how TAR may affect Brady compliance. Using original computer science simulations, we show that TAR can either hide or help to expose Brady evidence, depending on how it is configured. From these results we derive three recommendations: prosecutors should run TAR separately for inculpatory and exculpatory/impeachment evidence; courts should permit TAR coding of Brady material even when active searching is constitutionally contested; and procurement guidelines should favor flexible classifiers. More broadly, our examination of TAR highlights unresolved tensions in Brady doctrine: whether liability attaches when the prosecution possesses but does not know about exculpatory or impeachment evidence; how Brady interacts with Fourth Amendment privacy protections; and whether Brady should be limited to preventing suppression or expanded into a full duty to assist defense investigations.

As technology advances, the manipulation of digital content has become more sophisticated and accessible than ever. It is now easier to generate or alter photos, videos, audio recordings, and even handwriting, raising critical questions about the integrity of evidence presented in legal proceedings. With the rise of AI-generated content,  we must consider how these advancements can potentially compromise the  reliability of evidence in court.
This presentation addresses the need for legal professionals to understand and respond to the challenges posed by AI manipulation. We will explore whether our existing authentication rules are sufficient to withstand the complexities introduced by AI-generated evidence.

This panel discussion will cover the creation of Pennsylvania's AI & Emerging Technologies Committee, the committee's work, and the creation of a model AI Policy for PD offices to adapt and use.