The course is organized into five weekly modules, combining practical strategies, examples, and activities:

Introduction Module – Where We’ve Been, Where We’re Headed 

Three years after ChatGPT’s release, this module separates what’s actually changed in AI from what’s just gotten louder — the competitive and financial landscape among AI companies, and how technology companies have become energy companies. It’s the level set for the rest of the course: the conversation has to move from “which chatbot” to “is our data ready?”

Learning Outcomes:

• Separate what’s actually changed in AI since 2023 from what’s just gotten louder
• Understand the current competitive and financial landscape among AI companies — and why “who’s winning” is the wrong question
• Recognize how AI has become an energy and infrastructure story, not just a software one

Module 1: AI as infrastructure, for yourself and your newsroom

AI isn’t a technology you bolt onto your newsroom. It’s infrastructure — like your CMS, your analytics stack, your distribution.

Learning Outcomes:

• Distinguish between “using AI” (adding a chat tool) and “AI infrastructure” (reorganizing your data layer)
• Audit your newsroom’s information architecture — or your own, if you’re working solo
• Identify what “good data” looks like for journalism

Module 2: Building your personal Obsidian and MCP server

The fastest way to understand what AI infrastructure actually means is to build a small piece of it yourself. This module has you construct a personal “external brain” — then shows you the identical architecture at newsroom scale.

Learning Outcomes:

• Build a working personal knowledge vault and connect an AI assistant to it safely
• Understand what MCPs are and why they matter, from the inside, because you built one
• Map your newsroom’s knowledge repositories to potential MCP use cases
• Design a newsroom MCP strategy

Module 3: Agents and agentic workflows

Agentic AI isn’t a general, futuristic idea — it’s the next architecture, and journalism is better positioned to lead it than most industries realize, if newsrooms build the trust and verification layer in from the start rather than bolting it on after.

Learning Outcomes:

• Understand what makes an AI system “agentic,” and why journalism’s existing instincts (anticipation, source-vetting, distribution) translate directly
• Understand the trust and information-revenue questions agentic AI raises for the industry
• See a real, working example of “architecture as oversight” — a newsroom agent built so it cannot assert unverified facts
• Build the verification habits agentic systems require

Module 4: Technical literacy

AI fluency isn’t about being technical. It’s about knowing what AI can and cannot do on your beat.

Learning Outcomes:

• Understand LLM capabilities and limits (hallucination, knowledge cutoffs, bias in training data)
• Map AI use cases to specific beats, and know when not to use AI at all
• Build a beat-specific AI literacy and a personal policy for writing with AI

Module 5: Building a newsroom culture and strategy

A winning data set won’t matter in a newsroom that is resistant to change. This module provides a roadmap for who belongs at the table, how to track experiments, and turn five weeks of work into a strategy memo for leadership.

Learning Outcomes:

• Design an AI literacy and experimentation program for your newsroom
• Build governance structures that include journalists’ perspectives, not just the people buying the tools
• Build a practical roadmap for your newsroom’s AI strategy — next 90 days, 6 months, 12 months

 

Enroll now to get immediate access to the Introduction Module and begin building a practical AI strategy for your newsroom.

Questions? Contact us at journalismcourses@austin.utexas.edu.