\n<\/p>\n
Amazon Web Services\u2019 annual tech conference AWS re:Invent has wrapped up its first official day of programming and has already delivered an endless stream of product news. <\/p>\n
The unsurprising theme is AI for the enterprise, although this year it\u2019s all about upgrades that give its customers greater control to customize AI agents \u2014 including one that AWS claims can learn from you and then work independently for days.<\/p>\n
AWS re:Invent 2025<\/a>, which runs through December 5, started with a keynote from AWS CEO Matt Garman, who leaned into the idea that AI agents can unlock the \u201ctrue value\u201d of AI.<\/p>\n \u201cAI assistants are starting to give way to AI agents that can perform tasks and automate on your behalf,\u201d he said during the December 2 keynote. \u201cThis is where we\u2019re starting to see material business returns from your AI investments.\u201d<\/p>\n While AI agent news promises to be a persistent presence throughout AWS re:Invent 2025, there were other announcements, too. Here is a roundup of the announcements that got our attention. TechCrunch will continue to update this article through the end of AWS re:Invent, so be sure to check back.<\/p>\n AWS introduced a new version of its AI training chip<\/a> called Trainium3 along with an AI system called UltraServer that runs it. The TL;DR: This upgraded chip comes with some impressive specs, including a promise of up to 4x performance gains for both AI training and inference while lowering energy use by 40%.<\/p>\n AWS also provided a teaser. The company already has Trainium4 in development, which will be able to work with Nvidia\u2019s chips.<\/p>\n Techcrunch event<\/p>\n \n\t\t\t\t\t\t\t\t\tSan Francisco<\/span> AWS announced new features in its AgentCore AI agent building platform<\/a>. One feature of note is Policy in AgentCore, which gives developers the ability to more easily set boundaries for AI agents.<\/p>\n AWS also announced that agents will now be able to log and remember things about their users. Plus it announced that it will help its customers evaluate agents through 13 prebuilt evaluation systems.<\/p>\n AWS announced three new AI agents (there is that term again) called \u201cFrontier agents,\u201d including one called \u201cKiro autonomous agent\u201d that writes code and is designed to learn how a team likes to work so it can operate largely on its own for hours or days.<\/a><\/p>\n Another of these new agents handles security processes like code reviews, and the third does DevOps tasks such\u00a0as preventing incidents\u00a0when pushing new code live. Preview versions of the agents are available now.<\/p>\n AWS is rolling out four new AI models<\/a> within its Nova AI model family \u2014 three of which are text generating and one that can create text and images.<\/p>\n The company also announced a new service called Nova Forge that allows AWS cloud customers to access pre-trained, mid-trained, or post-trained models that they can then top off by training on their own proprietary data. AWS\u2019s big pitch is flexibility and customization.<\/p>\n The ride-hailing company was among many AWS customers that piped up during the event<\/a> to share their success stories and evidence of how products affected their business. Lyft is using Anthropic\u2019s Claude model via Amazon Bedrock to create an AI agent that handles driver and rider questions and issues. <\/p>\n The company said this AI agent has reduced average resolution time by 87%. Lyft also said it has seen a 70% increase in driver usage of the AI agent this year. <\/p>\nAn AI training chip and Nvidia compatibility<\/h2>\n
\n\t\t\t\t\t\t\t\t\t\t\t\t\t|<\/span>
\n\t\t\t\t\t\t\t\t\t\t\t\t\tOctober 13-15, 2026<\/span>\n\t\t\t\t\t\t\t<\/p>\n<\/p><\/div>\n<\/p><\/div>\n<\/div>\nExpanded AgentCore capabilities<\/h2>\n
A nonstop AI agent worker bee<\/h2>\n
New Nova models and services<\/h2>\n
Lyft\u2019s argument for AI agents<\/h2>\n
An AI Factory for the private data center<\/h2>\n