<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Stephen Ratner - Writing</title><link>https://www.stephenratner.co/</link><description>Writing by Stephen Ratner on AI, privacy, and product law.</description><language>en-us</language><atom:link href="https://www.stephenratner.co/feed.xml" rel="self" type="application/rss+xml"/>
<item><title>The Lawyer Sets the Limits</title><link>https://www.stephenratner.co/writing/agentic-practice-of-law.html</link><guid isPermaLink="true">https://www.stephenratner.co/writing/agentic-practice-of-law.html</guid><pubDate>Sun, 18 Oct 2026 12:00:00 -0400</pubDate><description>An AI agent can carry out a decision. The harder question is how it knows when that decision has stopped applying.</description></item>
<item><title>The Federal Framework</title><link>https://www.stephenratner.co/writing/federal-framework.html</link><guid isPermaLink="true">https://www.stephenratner.co/writing/federal-framework.html</guid><pubDate>Sun, 18 Oct 2026 12:00:00 -0400</pubDate><description>One national AI rulebook should give consumers more power and businesses a clearer account of what they owe.</description></item>
<item><title>The Patient Is Owed an Answer</title><link>https://www.stephenratner.co/writing/malpractice-and-agency.html</link><guid isPermaLink="true">https://www.stephenratner.co/writing/malpractice-and-agency.html</guid><pubDate>Sun, 18 Oct 2026 12:00:00 -0400</pubDate><description>Hospitals and developers can divide responsibility for medical AI without leaving the patient caught between them.</description></item>
</channel></rss>
