Summary: A rigorous methodological critique calls into question the scientific foundations of much contemporary consciousness research. The authors argue that widely used neuroscientific methods may be unable to distinguish subjective experience from general information processing, undermining many high-profile claims about sentience in non-human systems.
This analysis shows that common experimental paradigms can conflate conscious awareness with broader cognitive functions, which in turn may have amplified public and academic claims that animals, fetuses, lab-grown organoids, or advanced AI systems possess subjective experience.
Key Facts
- The Measurement Crisis: The central issue is not whether particular non-human systems are conscious, but whether current neuroscientific tools actually measure consciousness itself rather than general informational processing.
- The Methodological Loop: Common paradigms—such as visual masking, binocular rivalry, and threshold-based detection tasks—do more than alter reported awareness. They also change the brain’s capacity to integrate and process information, which can lead researchers to mistake reductions in information processing for a loss of subjective experience.
- The Non-Human Entity Boom: This ambiguity has practical consequences. Unclear measurement criteria have helped fuel bold claims about consciousness in animals, AI models, fetal development, and brain organoids—claims that may rest on markers sensitive to processing capacity rather than conscious experience.
- The Threat of Behaviorist Backlash: The authors warn that history offers a cautionary example. In the late 19th and early 20th centuries, poorly grounded assertions about consciousness contributed to a backlash that fostered behaviorism and delayed progress in the scientific study of subjective experience for decades.
- The Dissociation Roadmap: The paper highlights neuropsychological dissociations—such as blindsight and hemispatial neglect—where behavior and perception diverge from conscious awareness. These cases demonstrate that subjective experience and information processing can be biologically separable, suggesting clearer experimental pathways.
- High-Stakes Ethical Grounding: Senior author Hakwan Lau stresses that, because scientific claims about consciousness increasingly affect animal welfare, AI policy, and biomedical rules, researchers must pursue greater conceptual clarity and stricter methodology before translating findings into policy.
Source: Institute for Basic Science
As artificial intelligence advances, questions that were once philosophical are now urgent scientific and policy matters: Can AI be conscious? Do animals, organoids, or fetuses have subjective experience?
A team led by Director Hakwan Lau at the Center for Neuroscience Imaging Research (Institute for Basic Science), with collaborators from Université de Montréal and New York University, has published a critical analysis arguing that current empirical approaches may not be able to answer those questions reliably.
Rather than attempting to settle the status of specific systems, the paper scrutinizes whether the experimental markers commonly used in neuroscience genuinely index subjective awareness. The authors contend that many of those markers are sensitive to the brain’s overall information-processing ability, and therefore may not isolate consciousness per se.
“Many current theories of consciousness appear to be supported by a range of experimental findings,” Lau explains. “But those findings may actually reflect general information processing rather than consciousness itself—so it remains difficult to conclude that these theories truly explain consciousness.”
The critique emphasizes that paradigms such as masking, binocular rivalry, and perceptual thresholds do not only modify subjective reports. They also alter the underlying neural capacity for perception and cognition, producing potential confounds where task performance or neural signatures reflect processing limitations rather than the presence or absence of experience.
The authors warn this methodological confound has already shaped public debate and some scientific claims about consciousness beyond humans. Because many experimental “markers” track processing capacity, they may be poorly suited to justify claims about sentience in animals, AI, fetal stages, or cultured organoids.
Historical precedent is a key concern. The paper notes that earlier eras of psychology saw premature, inadequately supported assertions about mental life provoke a disciplinary collapse into behaviorism—an outcome that stunted the study of consciousness for decades. The authors argue that repeating that mistake would be costly both scientifically and ethically.
To advance the field, the paper recommends focusing on contexts where subjective awareness can be dissociated from behavior and perception. Clinical phenomena like blindsight and hemispatial neglect provide natural experiments showing that conscious experience and information processing can diverge. Studying such dissociations may yield more precise methods for testing consciousness.
Developing techniques capable of isolating subjective experience with greater precision will be essential before confidently applying neuroscientific findings to ethical and legal decisions about animals, AI systems, organoids, or fetuses. “If scientific claims about consciousness are going to influence discussions about animal welfare, AI ethics, or bioethics, then the scientific foundations supporting those claims must be especially rigorous,” Lau says.
The authors hope their analysis prompts the field toward stricter methodological standards, clearer conceptual definitions, and research designs that can better distinguish genuine subjective awareness from general information processing.
Key Questions Answered:
A: Current scientific tools often cannot separate neural or computational signatures of feeling from those of complex data processing. Experimental manipulations used to probe awareness frequently change how much information a brain or system can handle, producing signals that look like a change in consciousness but may actually reflect altered processing capacity.
A: Blindsight illustrates that visual information can guide behavior without entering conscious awareness. That dissociation demonstrates that processing and experience are distinct. By modeling and testing where these channels diverge, researchers can design experiments that more reliably test whether a system truly has subjective experience rather than just sophisticated information processing.
A: Premature or poorly supported claims can provoke a defensive backlash that undermines an entire field. A century ago, similar controversies contributed to the rise of behaviorism and long-lasting skepticism toward consciousness research. Today, with ethical and legal consequences at stake, methodological rigor is essential to avoid repeating that history.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this AI and consciousness research news
Author: William Suh
Source: Institute for Basic Science
Contact: William Suh – Institute for Basic Science
Image: The image is credited to Neuroscience News
Original Research: Open access. The Ethical Impasse of Current Consciousness Science by Vincent Taschereau-Dumouchel, Jun Seo Hwang, Hakwan Lau, and Joseph E. LeDoux. Neuron.
DOI: 10.1016/j.neuron.2026.04.007
Abstract
The Ethical Impasse of Current Consciousness Science
Growing attention to consciousness in animals, fetuses, organoids, and AI has produced strong claims based on experimental markers that may primarily reflect the capacity for information processing rather than subjective experience itself. As a result, the relevance of many such markers for adjudicating theories of consciousness may be limited. The paper argues for experimental and conceptual advances that can more cleanly isolate subjective awareness from general processing, to ensure that scientific conclusions and policy decisions rest on robust evidence.