<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Risk on SailingDataLakes</title><link>https://sailingdatalakes.com/tags/risk/</link><description>Recent content in Risk on SailingDataLakes</description><generator>Hugo -- gohugo.io</generator><language>en</language><lastBuildDate>Sun, 09 Aug 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://sailingdatalakes.com/tags/risk/index.xml" rel="self" type="application/rss+xml"/><item><title>In Claude We Trust? Accountability and Traceability in AI-Assisted Fraud Analysis</title><link>https://sailingdatalakes.com/posts/in-claude-we-trust/</link><pubDate>Sun, 09 Aug 2026 00:00:00 +0000</pubDate><guid>https://sailingdatalakes.com/posts/in-claude-we-trust/</guid><description>Everyone I work with uses Claude now — generating analyses, writing up docs, building the case for or against a decision. It&amp;rsquo;s made us faster. It&amp;rsquo;s also surfaced a set of problems that I think the industry, broadly, hasn&amp;rsquo;t caught up to yet.
I&amp;rsquo;ve noticed two flavors of this across teams I&amp;rsquo;ve worked with. More experienced people tend to present AI-assisted analysis with enough confidence that nobody thinks to double-check it.</description></item></channel></rss>