<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" 
     xmlns:content="http://purl.org/rss/1.0/modules/content/"
     xmlns:wfw="http://wellformedweb.org/CommentAPI/"
     xmlns:dc="http://purl.org/dc/elements/1.1/"
     xmlns:meneame="http://meneame.net/faq-es.php"
 >
<channel>
	<title>Menéame: comentarios [4168276]</title>
	<link>http://www.meneame.net</link>
	<image><title>www.meneame.net</title><link>http://www.meneame.net</link><url>http://cdn.mnmstatic.net/m/tecnología/img/mnm/eli-rss.png</url></image>
	<description>Sitio colaborativo de publicación y comunicación entre blogs</description>
	<pubDate>Wed, 15 Apr 2026 10:15:22 +0000</pubDate>
	<generator>http://blog.meneame.net/</generator>
	<language>es</language>
	<item>
		<meneame:comment_id>44684586</meneame:comment_id>
		<meneame:link_id>4168276</meneame:link_id>
		<meneame:order>1</meneame:order>
		<meneame:user>Mendax</meneame:user>
		<meneame:votes>0</meneame:votes>
		<meneame:karma>7</meneame:karma>
		<meneame:url>https://www.meneame.net/m/tecnología/openai-tac-acceso-ia-defensa-cibernetica</meneame:url>
		<title>#1 OpenAI TAC: acceso a IA para defensa cibernética</title>
		<link>https://www.meneame.net/m/tecnología/openai-tac-acceso-ia-defensa-cibernetica/c01#c-1</link>
		<pubDate>Wed, 15 Apr 2026 10:15:22 +0000</pubDate>
		<dc:creator>Mendax</dc:creator>
		<guid>https://www.meneame.net/m/tecnología/openai-tac-acceso-ia-defensa-cibernetica/c01#c-1</guid>
		<description><![CDATA[<p>Parece un movimiento de OpenAI para enfrentar todo el ruido que ha hecho Claude Mithos estos últimos días. De hecho en el post original (que no me ha dejado mandar por tener protección antibots) [ <a href="https://openai.com/index/scaling-trusted-access-for-cyber-defense/" title="openai.com/index/scaling-trusted-access-for-cyber-defense/" rel="nofollow">openai.com/index/scaling-trusted-access-for-cyber-defense/</a> ] escriben algo que parece dirigido a Anthropic, que de momento ha decido mantener restringido su modelo: <i>&#34;Democratized access: Our goal is to make these tools as widely available as possible while preventing misuse. We design mechanisms which avoid arbitrarily deciding who gets access for legitimate use and who doesn’t. &#34;</i>.  Mithos ataca, GPT-5.4-Cyber defiende.</p><p>&#187;&nbsp;autor: <strong>Mendax</strong></p>]]></description>
	</item>

</channel>
</rss>
