Press releases aren't dead-they're feeding AI

Press Releases Aren't Dead - They've Become Training Data

People keep asking if PR is outdated. If press releases are noise. The short answer: no. In the AI era, a good release feeds two audiences at once-journalists and large language models.

The flow looks like this: release goes out, a reporter turns it into a story, and AI answers pull from that reporting (and sometimes the release itself). Your message moves from your wire to headlines to AI summaries that millions read. That's leverage.

Why AI likes press releases

• They're structured: clear headline, subhead, lead, quotes, boilerplate, contacts.
• They're factual: names, dates, numbers, locations-easy for models to parse and cite.
• They're concise: big picture first, then supporting details. Minimal filler.

That structure helps journalists work fast. It also makes it simple for LLMs to extract the "who, what, where, when, why" without guessing.

From Wire to Headline to AI Answer

Most AI answers you see are stitched from reported pieces. Those pieces often start with your release. Sometimes AI will surface newswire copy directly if it's the clearest source.

Not all models behave the same. Some versions of ChatGPT may not browse by default, while others and tools like Gemini can pull fresh content. Either way, millions will encounter content that began as a release. If the facts are yours, you're in the conversation.

A quick history note

The first modern press release dates back to 1906 and Ivy Lee's statement about a railroad accident. Over a century later, the format still works because it centers facts and clarity. That same clarity plays well with algorithms and editors alike.

Read more about the first press release.

How to write releases that work for humans and AI

Make the headline do the heavy lifting

• Be specific: company, action, outcome. No fluff, no hype.
• Front-load the news: if someone only reads the headline, they should still get the point.
• Keep it under ~12 words when possible. Clear > clever.

Use a subhead to add essential context

• Who it affects, where it applies, and why it matters.
• Introduce one key stat, date, or differentiator you'll expand on below.

Write a lead anyone can quote

• In 2-3 sentences, answer: what's new, for whom, when, and where.
• Include one concrete metric: revenue, users, funding, time saved, % change.

Keep paragraphs short and skimmable

• 2-4 sentences per paragraph.
• Use bullets for lists of features, partners, locations, or dates.
• Spell out acronyms on first use and name all entities precisely.

Quotes that add meaning (not adjectives)

• Quote a decision-maker with title and full name.
• Use quotes to explain the "why," impact, or next step-not to restate the news.

Dates, times, and numbers

• Use an exact date with time zone.
• Format numbers consistently and avoid vague ranges.
• If you mention money, note currency.

Proof and sources

• Link to a public page with more detail (product page, FAQ, research PDF).
• Add one chart or image with a clear caption to anchor the story.

Boilerplate and contacts

• Boilerplate: one tight paragraph with what the company does, for whom, and scale (customers, markets, or key metric).
• Media contact with name, email, and phone. Add a backup contact for events or after-hours.

Distribution still matters

A release no one sees won't feed headlines-or AI. Use a wire when the news merits broad reach, but don't stop there. Pitch short, tailored notes to the right reporters with the angle that fits their beat.

Publish on your newsroom first with a clean URL and linkable assets. Share a brief version on LinkedIn with the core stat and a visual. Align the publish time with your audience's time zone, not yours.

Checklist: make your release AI-friendly

• Descriptive headline + clear subhead
• Lead with the "what, where, when, why it matters"
• Short paragraphs and bullet lists for key details
• Named entities spelled consistently throughout
• One strong metric and one strong quote
• Image or chart with caption
• Links to deeper sources
• Boilerplate + contact info

What to measure beyond pickups

Track the usual: placements, unique visitors, share of voice, referral traffic. Then add a new layer: how your message shows up in AI answers over time.

• Run periodic queries your buyers would ask and see whose facts get cited.
• Note which phrasing sticks-headlines and subheads often become the snippet.
• Log where quotes appear and which outlets tend to "seed" AI responses.

Common mistakes to avoid

• Burying the lede under three paragraphs of corporate speak
• Hype without numbers
• Missing location, date, or availability details
• Paragraphs longer than four sentences
• Quotes that say nothing ("we're thrilled…")

The opportunity for PR and Comms

Press releases haven't faded. They've gained a second life as inputs for how people now search and learn. If your release is tight, credible, and easy to parse, it helps reporters work faster and gives AI fewer chances to guess.

That's good for accuracy, good for reach, and good for your brand. Keep the facts clean, the structure simple, and the story useful. The medium still works-as long as the message earns its place.

Upskill your team for AI-era PR

If you want practical training on AI tools and workflows for communications work, explore curated options by role here: Complete AI Training - courses by job.