AI Briefing#hallucination
The latest AI and developer news about #hallucination, with the original source and a short summary.
The latest AI and developer news about #hallucination, with the original source and a short summary.
Banksalad has filed three patents for technologies that control errors in financial transactions by AI agents.
Qwen attached citations to 98.3% of answers and was the only engine to ignore a planted fake concept, outperforming ChatGPT and DeepSeek in source reliability.
A global initiative with over 100 contributors from 50 countries found that AI models often provide authoritative but incorrect answers to culturally specific riddles.
A large-scale audit of 160,860 citations reveals that nearly 40% of displayed source links are never referenced in the final AI-generated answer.
An analyst at U.S. Special Operations Command Pacific (SOCOM Pacific) prepared to intercept Chinese ships based on AI-generated false intelligence, but the operation was canceled just before execution when the report was revealed to be fabricated.
34.7% of Perplexity's numerical citations were found to link to pages that either do not contain the cited figures or are inaccessible.
The LangChain team presented methods to address hallucination and context issues when generating SQL queries with LLMs.
The study comparatively analyzed preference alignment methods and datasets for multimodal LLMs and proposed BDHS.
A study shows that existing metrics for chest X-ray report generation overlook clinical term deletion and bias.
RL-finetuned VLMs show improved visual reasoning performance, but exhibit vulnerability to textual perturbations.
A new benchmark, OpenBioRQ, points out that 15.9% of papers cited by AI agents do not match the content of their claims.
It was revealed that the hallucination phenomenon in which the multimodal model Inter-1 repeatedly outputs a specific phrase on silent videos is caused by a combination of prompt examples and post-training.
It has been revealed that each LLM model and version has a unique bias toward repeatedly generating specific person names.
An Ontario audit found that AI scribe systems for doctors showed serious hallucinations, including errors in medication information and missing patient information.
A new evaluation methodology has been released addressing the 'source-boundary' problem, where LLMs mistake information within context.
Explicitly stating source boundaries was needed for LLMs to distinguish text usable as grounds for an answer.
A Maine lawyer was sanctioned after AI errors turned up in court filings.
It argues that reducing hallucination requires metacognition and faithful expression of uncertainty.
OpenAI has updated ChatGPT's default model to GPT-5.5 Instant, improving accuracy and personalization.
SOB unveiled a new benchmark that verifies not just JSON schema pass rate but also value accuracy.
SOB has released a new benchmark that measures not only JSON parsing but also value accuracy.
This introduces the causes of LLMs speaking with confident tones even when hallucinating, and MIT's research findings to address this.
Across a 90-day tracking of 42 recurring tasks, Opus 4.6's hallucination rate surged.
While actively promoting Copilot, Microsoft has cautioned users through its terms of service, explicitly noting the AI's potential for errors and placing responsibility on the user.