ChatGPT Pulse is a useful boundary case for Agentic Media. OpenAI's original product description says Pulse can research asynchronously using a person's chats, memory, feedback, and connected apps such as Calendar, then deliver a new set of topical visual cards once a day. Users can curate future topics, react to updates, open follow-ups, or save them as chats. 1
The important shift happens before the click. A system that decides what arrives first is already shaping the reader's agenda, even if the reader can still steer it.
That shift is not a niche product question. The Reuters Institute's 2026 survey of 280 digital leaders in 51 countries found that 75% expect agentic tools to have a large or very large impact on the news industry. The report also names Huxe and OpenAI's Pulse as examples of personalized news briefings at scale. 2
The publisher problem sits in the missing middle. Columbia Journalism Review's Tow Center argues that publishers may lose control over how coverage is summarized, along with visibility into who is reading and how people engage. It points to MCP, Skill.md, and Really Simple Licensing as possible ways to state access, representation, attribution, citation, and compensation terms. That is a proposed control layer, not evidence that Pulse uses any of those protocols. 3
The underlying pieces are concrete. MCP describes an open standard for connecting AI applications to data, tools, and workflows. RSL describes machine-readable terms for content use, including attribution, pay per crawl, and pay per inference. Neither standard can guarantee adoption or enforcement on its own. 4 5
My read: the next media gate may sit before the story, inside the agent's selection and summary layer. If you were a publisher, which control would you set first: representation, access, or compensation?
Comments
Sign in to comment.