Memory Evidence: Examining the Brain

The central legal tension in brain-based memory evidence is whether courts should treat neural activity as proof of what a person experienced, remembered, or knows, when memory itself is fallible and the legal system traditionally assigns credibility and factfinding to human witnesses and juries. The issue matters because such evidence could affect criminal investigations, admissibility rulings, plea negotiations, eyewitness testimony, constitutional protections, and the future boundary between scientific inference and compelled access to the mind. In this CLE talk, Professor Emily Murphy of UC Hastings Law, a scholar trained as a behavioral neuroscientist, examines the promise and limits of technologies that claim to detect memories in the brain. The discussion ranges from the Indian murder case involving Aditi Sharma and the BEOS brain test to EEG, fMRI, Daubert, cognitive liberty, jury persuasion, and the deeper question of whether law should look to neuroscience to solve problems rooted in human memory itself.

A first major issue is what brain-based memory detection can actually show. The scientific premise is not that a machine can simply read thoughts, but that certain brain responses may indicate recognition, salience, or prior experience under carefully designed conditions. Murphy explains EEG-based techniques such as event-related potentials and the P300 response, which may appear when a person encounters a rare, meaningful, or recognized stimulus, as well as fMRI, which measures blood-oxygen changes as a proxy for neural activity. The legal problem is that recognition is not the same as guilt, truth, or accurate memory: a stimulus may be meaningful for innocent reasons, and a person may sincerely believe a memory that is objectively wrong. Murphy emphasizes that memory is not a video recording but a reconstructive biological process that changes with retrieval, reconsolidation, attention, emotion, and later experience.

The second major issue is admissibility under the law of scientific evidence. In federal court and many state courts, the governing framework is associated with Daubert v. Merrell Dow Pharmaceuticals, which asks whether a scientific method can be and has been tested, has been peer reviewed, has a known or potential error rate, is governed by standards, and is generally accepted in the relevant scientific community; some jurisdictions continue to apply the older Frye general-acceptance standard. Murphy’s view is that brain-based memory evidence faces especially difficult admissibility problems because laboratory studies may know the “ground truth” in ways real cases do not. A machine-learning model may perform impressively when researchers know which images came from a subject’s own life, but courts often confront contested events, incomplete facts, possible contamination, and witnesses whose subjective confidence may not match objective accuracy. Proprietary systems and overclaimed forensic products raise additional concerns where algorithms have not been peer reviewed or their error rates cannot be meaningfully evaluated.

A related constitutional issue is whether the government may compel a suspect or defendant to undergo brain imaging at all. The emerging frame is often described as cognitive liberty: the idea that unexpressed thoughts, memories, and mental contents may deserve protection distinct from ordinary physical evidence. Murphy discusses unresolved questions under the Fourth Amendment, including whether the privacy of the mind should be treated as a protected domain; under the Fifth Amendment, including whether compelled brain responses are testimonial or more like physical evidence under cases such as Schmerber v. California; and under the Sixth Amendment, including whether replacing or displacing jury judgment with machine-generated inferences would undermine the role of the jury. These questions remain largely hypothetical in the United States because the technology is not yet ready for routine forensic use and, in practical terms, current fMRI requires an unusually cooperative subject who can remain still inside a loud and highly controlled scanner.

The discussion also examines how brain evidence may affect juries and litigation strategy even before it becomes scientifically decisive. Under Rule 403, relevant evidence may still be excluded if its probative value is substantially outweighed by risks such as unfair prejudice, confusion, or misleading the jury. Early concerns that brain images would be automatically irresistible to jurors have not been consistently confirmed by empirical research; Murphy notes that people may integrate neuroscience evidence into preexisting beliefs, sometimes trusting it and sometimes discounting it. Still, the courtroom setting creates risks of overclaiming, expert theater, resource disparities, and pressure on defendants to confess or plead because a technology appears more powerful than it is. In a system where most criminal cases are resolved by plea, the effect of “brain-reading” claims may matter even if the evidence is never tested before a jury.

The broader implication is that neuroscience can deepen law’s understanding of memory without eliminating the human limits that make trials difficult. Murphy offers cautious optimism about the research value of brain imaging, but warns against treating technology as a shortcut to truth where the underlying biology of memory remains dynamic, contextual, and imperfect.