Google DeepMind researchers introduced SynthID-Text, a production-ready text watermarking scheme for large language models that preserves text quality and enables high detection accuracy with minimal …
Anthropic, the firm behind the AI model Claude, has announced that text generated by any models launched on or after 2 August will be invisibly watermarked to indicate AI authorship, with images carry…
A Perspective in Nature Reviews Drug Discovery finds that despite a decade of intense interest, AI in drug discovery has shown disappointingly limited clinically relevant impact, with no evidence yet …
A study published in Nature found that government control of media influences large language models (LLMs) through their training data, with models showing stronger pro-government responses in languag…
A new study in npj Climate Action models AI as a bidirectional productivity amplifier in a global computable general equilibrium model, finding that under parallel adoption scenarios, net annual CO₂ e…
A study posted on arXiv in late July found that AI systems are not yet capable of fully automating open-ended research, with an AI system built by Princeton University researchers scoring 2/6 and 1/6 …
A 2024 Nature paper finds that 'open' AI systems often lack precision in their claims of openness, and that openness alone does not reduce industry concentration in AI development. The authors, examin…
QED Science, a Tel Aviv-based start-up, has developed an AI tool that ranks preprints by originality and validity, and its analysis of 57,455 bioRxiv preprints from May 2025 to April 2026 selected a '…
A News and Views article in Nature (656, 42-44, 2026) highlights that privacy risks from medical AI tools are not equally shared, citing a study by Knolle et al. in Nature 656, 192-198 (2026). The art…
A study of 971 rural teachers in Guangxi, China, found that technical support was the strongest factor associated with attitudes toward generative AI adoption (β = 0.515, p < 0.01), followed by percei…
A multi-ancestry genome-wide association study meta-analysis of 2,563,755 individuals (54,629 cases and 2,509,126 controls) from 11 cohorts identified 26 risk loci for fibromyalgia, with the strongest…
A double neural bypass (DNB) neuroprosthetic system restored hand movement and sensation in a participant with chronic C4 sensory/C5 motor complete tetraplegia, enabling functional abilities such as s…
An artificial-intelligence tool called Raygun can shrink or enlarge natural proteins without disrupting their shape or function, according to a study published today in Nature. Developed by Rohit Sing…
Researchers led by E. Goh and colleagues at Google, Meta, Amazon, Anthropic, OpenAI, and Microsoft argue in Nature Medicine that existing benchmarks for medical AI are misleading and insufficient, urg…
A new 'humanizer' tool released on 20 June by Jie Ding, a machine-learning researcher at the University of Minnesota, can erase signs of AI-written text in papers and grant proposals. While some resea…
Nobel Prize-winning chemist Omar Yaghi has left the United States to lead a new artificial-intelligence-assisted materials discovery institute at Tsinghua University in Beijing, China, citing cuts to …
Researchers built an archive of 70 preregistered survey experiments and found that GPT-4 can predict the outcomes of social science experiments with accuracy similar to pooled human forecasts, though …
Universities are increasingly relying on AI-detection software to identify AI-generated text in student applications and assignments, but the tools are unreliable and can falsely flag human-written co…
Researchers introduced a conceptual framework for 'LLM Pollution' in online behavioural research, identifying three variants—Partial LLM Mediation, Full LLM Delegation, and LLM Spillover—that distort …
Generative artificial intelligence has learned the structure of the human palate from recipe data to create novel burgers optimized for taste, sustainability, or nutrition. In a blinded sensory evalua…