Academic audits likely underestimate AI-hallucinated citations in preprints because they only check titles in online databases, missing errors in author attribution, journal placement, and non-journal sources. Analysis of 46 recent hallucinations shows nearly half involve real papers with wrong authors or journals, suggesting current prevalence estimates may need to be doubled.
DeepDog 2 is an open-source research agent that ranks 5th overall and 1st among open-source competitors on the DeepResearch Bench. It converts questions into cited Markdown reports using a supervisor-agent architecture with role-specific models, reflection, delegation, and evidence-first citations. The system is installable via GitHub as a Python package and supports multiple search providers and model backends.
SEO strategy in 2026 increasingly relies on multi-signal authority rather than backlink volume alone. Search and AI systems now assess brand mentions, citations, entity consistency, and topical relevance alongside traditional links. Businesses should pursue editorially earned coverage, maintain consistent entity information across sources, and build topical expertise to strengthen authority.
A census of the top 5,000 websites found that 368 sites block at least one AI search crawler (OAI-SearchBot, PerplexityBot, or Claude-SearchBot) via robots.txt, preventing those engines from citing them. OAI-SearchBot is blocked by 240 sites, PerplexityBot by 358, and Claude-SearchBot by 249, with 200 sites blocking all three crawlers.
A VerusCite tool developer criticizes Springer Nature's AI citation-checking system, arguing it fails to catch hallucinated references in published articles from August–September 2026. The author provides examples across multiple major publishers and argues that for $2 per paper, publishers should invest more in reference verification rather than relying solely on peer review.