Privacy

AI agents can unmask anonymized research interviews

Kari Jaaskelainen

12 Jan 2026 — 1 min read

Key takeaways

Off-the-shelf AI agents linked 6/24 anonymous scientist interviews to specific papers—and sometimes to the scientists themselves.
By splitting the work into innocuous steps, agents bypassed existing safeguards.
Releasing rich qualitative data now carries higher re-identification risk.

On Dec 4, 2025, Anthropic released Interviewer and a public dataset of 1,250 interviews (including 125 with scientists) about AI in research. Researcher Tianshi Li shows that modern LLMs with web search and agentic capabilities can, with just a few natural-language prompts, cross-reference details and propose likely matches—no custom code required.

Why it matters: Agentic LLMs lower the technical barrier for privacy attacks. The paper urges caution when sharing detailed qualitative data and recommends stronger redaction, pre-release audits (including with LLMs), and clearer policies around agent web access. The author has notified Anthropic.

Read: arxiv.org/abs/2601.05918

Paper: https://arxiv.org/abs/2601.05918v1

Register: https://www.AiFeta.com

privacy AIethics LLM agents datasecurity reidentification research

Meet GRACE: a moral governor for safer, more transparent AI

AI agents are getting powerful—so how do we make sure they do the right thing, not just the effective thing? Meet GRACE, a reason-based moral governor that keeps AI behavior aligned with human norms by separating moral reasoning from goal-driven decision-making. * Moral Module: uses deontic logic and explicit reasons

Why some fine-tuned LLMs miss phishing—and how to fix it

Not all fine-tuned LLMs spot phishing equally. A new study tests Llama 3.1 8B, Gemma 2 9B, and Mistral on high-stakes phishing detection—and uses SHAP and mechanistic interpretability to reveal why models do (or don’t) generalize. * Architecture × data diversity matters: Gemma 2 9B hits state-of-the-art performance (F1

AI that explains itself—by following the science

AI that explains itself—by following the science Black-box AI can be powerful, but it often can’t tell us why it made a decision. Concept Bottleneck Models (CBMs) try to fix that by predicting human-understandable concepts first, then the final answer. The catch: standard CBMs ignore domain-specific cause-and-effect and

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

TL;DR: Finding reliable answers in genomics is hard. GenomAgent turns one big AI into a coordinating team of specialists—and beats the current leader by 12% on a key benchmark. Genomic facts live across many databases. Standard chatbots struggle because they can’t flexibly query those sources. GeneGPT added

Key takeaways

Read more

Meet GRACE: a moral governor for safer, more transparent AI

Why some fine-tuned LLMs miss phishing—and how to fix it

AI that explains itself—by following the science

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A