The press release used to have one job: help a journalist decide if a story is worth covering. That job still matters, but now it’s not the only one.
Today a press release can also shape how AI-powered tools such as ChatGPT, Google Gemini, Claude, and Perplexity understand a company, summarize an issue or compare organizations months after a release is issued. These platforms are becoming a first stop for research, helping users gather information, compare companies, and understand issues before they ever visit a website or speak with a representative.
According to Muck Rack’s May 2026 What Is AI Reading? study, which analyzed more than 25 million links cited by AI platforms, approximately 99% of AI citations come from non-paid sources. Press releases accounted for 1.1% of citations across ChatGPT, Claude, and Gemini, compared with 0.3% for paid content.
That shift changes the role of the press release. It is no longer only a vehicle for media pickup or a searchable announcement on a company newsroom. It is also a source document that can influence how AI systems understand, summarize, and surface information about an organization.
The takeaway is not that every release will show up in an AI-generated answer. Let’s not get carried away. The takeaway is that press releases are now part of the broader information ecosystem AI systems use to interpret organizations, products, executives and issues.
For communicators, this creates a new responsibility: announcements must be written not only to earn attention, but also to preserve accuracy. They need to be newsworthy for reporters, clear for readers, and structured enough that AI tools can understand them without flattening the meaning.
How AI Changes the Way Press Releases Work
A journalist and an AI system may read the same release – but for different reasons.
A journalist is evaluating whether there is a timely, relevant story worth pursuing. An AI system is identifying facts, entities, relationships, and context it can retrieve later. That difference does not require communicators to abandon the fundamentals of strong release writing. It does require more intention around how information is structured.
First, AI systems benefit from a clear hierarchy. Descriptive headlines, clear opening paragraphs, and well-labeled sections help establish what matters most. A release that is easy for a reporter to scan is also easier for a machine to interpret. Clear is not boring. Clear is useful.
Second, the most important information needs to be easy to extract for AI systems. The first paragraph should quickly answer who, what, and why. Quotes should communicate a clear point of view and be able to stand on their own. Clever framing can still have a place, but buried news creates risk when content is being summarized outside its original context.
Third, consistency matters over time. AI systems look for patterns. One announcement can introduce a topic. Consistent language across multiple announcements can help reinforce what an organization is known for. Linking to related releases, using consistent terminology and building thematic threads all help create a clearer body of information around a company’s priorities. This is where a press release stops being a one-off announcement and becomes part of the organization’s larger message architecture.
Finally, supporting elements matter. Images, captions, alt text and file names have long supported accessibility and search. In an AI-influenced discovery environment, they also provide additional context that can reinforce the message of the release.
What This Means for Communicators
The rise of AI discovery does not mean organizations need two separate press releases: one for reporters and one for machines. The better model is a dual-purpose press release.
A strong release still needs news value, context, and relevance for human audiences. It also needs structure, clarity, and consistency for AI systems that may summarize or retrieve it later.
Before issuing a release, communications teams should ask:
- Does the lead clearly explain what happened and why it matters?
- Could the quote stand alone and still communicate a meaningful point of view?
- Are we using language that connects to our broader message platform?
- Have we linked to related context that helps explain the announcement?
- Are supporting assets labeled and described in a way that reinforces the story?
These are not tricks for gaming AI. They are good communications habits made more important by a changing discovery environment.
In that sense, AI optimization is not a reinvention of press release writing. It is a sharper application of the same principles that have always mattered: credible information, objective language, clear organization, and purposeful messaging.
The difference is the environment those principles now operate in.
Today’s press release is entering an ecosystem where journalists, search engines, and AI systems are all listening. Communications teams that understand this shift will be better positioned to ensure their announcements are not only seen but accurately understood.
The goal is not to write for machines at the expense of people. The goal is to write clearly enough that both can understand the point.
And in a communications environment where being misunderstood can happen faster than ever, that clarity is not a nice-to-have. It’s the job.