A university president outlines six principles meant to help institutions resist political pressure on research funding, peer review and academic autonomy, citing examples from the United States and Hungary. The piece traces how funding cuts, shifting grant priorities and legal campaigns against institutions like Central European University have already eroded independent scholarship in some countries.
Writer Gwern observes that while fraudulent videos, photos, and Word documents are common online scams, editing existing PDFs to forge fake claims is essentially unheard of. He notes that searching for and downloading a paper's PDF almost always yields exactly what was originally published, aside from minor draft-versus-final discrepancies.
The US National Science Foundation will issue its fewest new grants in four decades after withholding roughly $1 billion for the White House's Grand Research Challenges initiative and pulling back recommendations for over 150 peer-reviewed proposals. Meanwhile, the National Natural Science Foundation of China is increasing its prestigious early-career grants by 50% this year, moving in the opposite direction as Washington retreats from research investment.
A commentary examines how generative AI tools, now used to summarize literature, draft text and respond to peer reviewers, are complicating traditional notions of who counts as an author on a scientific paper. The piece argues that authorship has always been a socially constructed status tied to accountability and trust, not just credit, and that AI's growing role in writing, review and publishing workflows demands deliberate public debate rather than decisions based on convenience or cost.
A Nature opinion piece argues that scientific publishing's peer-review system is strained by rising submissions, more complex manuscripts and a shortage of willing, timely reviewers. While most journals currently prohibit reviewers from using AI tools over confidentiality and accountability concerns, the author contends these blanket bans are difficult to enforce and only partially solve the underlying problems.