Philosophical Analysis: Large Language Models Lack True Agency and Moral Responsibility
A preprint posted to arXiv contends that large language models lack the genuine agency required for moral responsibility, distinguishing probabilistic output generation from authentic choice. The authors argue that LLMs operate through learned input-output mappings and exhibit only 'derived' rather than 'intrinsic' intentionality, meaning their outputs are not owned as commitments or guided by reasons. The paper matters because it directly challenges growing claims in AI discourse that sufficiently advanced LLMs may qualify as moral agents deserving of ethical consideration or accountability.
A paper submitted to arXiv on June 11, 2026, by Joseph Keshet and co-authors argues that attributing agency or moral responsibility to large language models is philosophically misguided. The authors contend that genuine moral responsibility requires commitment-bearing agency rooted in intrinsic intentionality and self-attributed action — properties they say LLMs fundamentally lack. While LLMs can produce coherent, normatively evaluable, and even morally reasoned outputs, the paper maintains this reflects probabilistic pattern-matching over training data rather than authentic deliberation. Crucially, the authors argue that stochastic sampling — the randomness introduced during text generation — does not constitute choice or authorship in any morally relevant sense. The paper systematically addresses counterarguments from philosophical positions including the intentional stance, functionalism, and compatibilism, concluding that none are sufficient to establish genuine LLM agency. The work contributes to an active debate in AI ethics about where moral and legal accountability should be located as AI systems become more capable.
What's missing
The paper is a preprint and has not yet undergone peer review, so its philosophical arguments have not been formally evaluated by the academic community. The abstract does not specify whether the authors engage with recent empirical work on LLM internals (e.g., mechanistic interpretability) that some researchers argue bears on questions of intentionality, nor does it address potential distinctions between current LLMs and future architectures that might incorporate more explicit goal-directed mechanisms.
What different sources said
- arXiv cs.AICenter
Why Sampling Is Not Choosing: Intentionality, Agency, and Moral Responsibility in Large Language Models
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.