OpenAI Academy rolls out AI training videos for journalists

The ChatGPT 101 for Journalists collection covers newsroom briefs, prompting, reusable Skills and agent workflows for reporters, editors, producers, newsroom leaders and educators.

Editorial image showing a person working on a laptop in a blue-lit digital workspace, used to illustrate OpenAI Academy’s new ChatGPT 101 for Journalists training videos for newsroom teams, editors, producers and educators.

OpenAI Academy has added ChatGPT 101 for Journalists videos covering newsroom prompts, Skills and agent workflows.

OpenAI has added a new set of Academy videos for journalists and news organizations, expanding its media-focused training with practical lessons on using ChatGPT in daily newsroom work.

The ChatGPT 101 for Journalists collection sits inside the OpenAI Academy for News Organizations, a learning community for editorial, business, product and technology teams in news media. The community is listed with 1,409 members and includes trainings, case studies and playbooks focused on how AI can be used in ways that align with newsroom values and audiences.

OpenAI Global Affairs announced the videos on LinkedIn, describing them as “specifically tailored for news organizations and journalists.”

The post said the tutorials are aimed at “reporters, editors, producers, newsroom leaders, and educators who want practical ways to start using AI in their daily work, without needing a technical background.”

The four-video collection includes Build Your Own Assignment Editor, Power Prompting, Upskill With Skills, and Skills vs. Agents. The videos were published between July 7 and July 13, 2026, and are presented by Evan Hirsch.

Training starts with daily newsroom work

The videos focus on newsroom routines rather than general AI awareness. Build Your Own Assignment Editor shows how a Workspace Agent can review approved sources such as a calendar, inbox, Slack or Teams channel, and documents, then turn that information into a short morning brief.

OpenAI Academy describes the use case as a way to help journalists organize deadlines, meetings, messages and focus time at the start of the day. The workflow is framed around triage, with the AI supporting planning rather than taking action on its own.

The video asks users to connect only approved apps, set limits around what the agent should use, and define what the agent should not do. The training also keeps human editorial judgment in the workflow, particularly around what to prioritize and what requires follow-up.

Power Prompting takes a familiar newsroom discipline and applies it to ChatGPT. The video uses a city council agenda item to show how a weak prompt can be rebuilt into a clearer reporting instruction. The structure covers role, task, context, output, limits and check, giving journalists a way to make ChatGPT responses more specific and easier to verify.

The lesson also covers how journalists can ask ChatGPT to identify uncertainty, missing facts and possible next reporting steps.

Skills and agents get newsroom guardrails

The collection then moves from one-off prompting into repeatable newsroom workflows.

Upskill With Skills explains how a ChatGPT Skill can capture a recurring process and return a consistent type of output each time. OpenAI gives examples including agenda scans, public record reviews, source follow-up lists and editor prep packets.

The video asks users to start with a ChatGPT account, a repeatable newsroom task and a sample input, such as an agenda, meeting packet or public document. It also covers how to build human verification into the workflow.

Skills vs. Agents draws a distinction between the reusable process and the workflow that can apply it. OpenAI Academy describes a Skill as the template for how a newsroom wants a brief structured, which sources should be cited, which claims need review and what the output should look like. An Agent is described as the workflow that can use approved tools, gather material, follow instructions and apply the Skill.

The video covers how to design an Agent with a clear beat, job and limits. It also emphasizes preserving source links, dates and human review, which are the controls most relevant to newsroom use.

OpenAI builds out news organization resources

The OpenAI Academy for News Organizations section also includes media case studies and workflow examples from publishers in several markets.

The examples listed include India Today, Mediengruppe Pressedruck, General-Anzeiger, Hearst Newspapers, The Quint, Deccan Herald, VG, DMG Media and El Comercio.

The use cases cover AI-powered audience forecasting, press release and police report processing, newsroom adoption, sales tools, reader assistants, instant infographics, journalist copilots and election guide workflows.

That wider collection gives the journalist training a more applied setting. The ChatGPT 101 videos teach basic workflow design, while the case studies show how news organizations are experimenting with AI across editorial, audience, commercial and operational tasks.

The most editorially significant part of the new training is its emphasis on limits. The videos repeatedly return to approved inputs, source links, dates, workflow boundaries and human review. Those are the areas where newsroom AI can quickly become risky if teams treat generative AI as a generic productivity shortcut rather than a process requiring editorial control.

OpenAI Global Affairs framed the lessons around giving journalists more time for work only people can do, including “asking better questions, checking the facts, making hard editorial calls, and serving their communities with care.”

The ChatGPT 101 for Journalists collection is live now in the OpenAI Academy for News Organizations community.

Previous
Previous

AI economy statement urges action on jobs, guardrails and institutions

Next
Next

Anthropic opens Claude for Teachers free to U.S. K-12 educators