AI for Communications: Augment vs. Supplement, the Intent Gap and AI-Ready Press Releases

Day 1 of ALI Conferences’ AI for Communications Virtual Certificate Series on August 18 opened with a quick poll. Most attendees said they use AI daily and that it adds value, but very few felt they were using it as effectively as they could. The rest of the day was built to close that gap.

Chairperson McClain Houston, Employee Experience and Communications Manager at AArete, admitted this was the first conference she’s chaired where she didn’t feel like an expert, and suggested nobody should claim to be one right now. Here are the frameworks and data communicators took away.

Augment or supplement: a decision framework for using AI at work

Houston’s organization rolled out Copilot with a goal of 100% adoption in six months, but little guidance on what to use it for. Her workshop offered a practical way to decide.

• Augment what you’re already good at. The question isn’t how AI can reduce your workload, but how it can improve your judgment.
• Supplement your gaps and the low-value, repeatable tasks that eat your week. For Houston, that means spreadsheets and weekly timesheets. The test: can AI do this responsibly without changing the quality or meaning of the work?

Before using AI on a task, run through five checks:

1. Judgment: does it need professional judgment? If yes, augment. If not, supplement.
2. Risk: if the task touches reputation, certifications, security or regulated data, keep AI out of it.
3. Context: mergers, layoffs and leadership changes carry layers AI can’t see. Draft with AI if you like, but apply human judgment about the politics.
4. Repeatability: predictable, recurring tasks are ideal to hand off.
5. Value: is this a good use of your time at all?

Ways communicators use AI to augment their strengths

• Role-play a message as an intern and as a managing director would read it.
• Flag jargon, acronyms and assumptions (the “curse of knowledge”).
• Ask what questions a strong draft will trigger before you send it.
• Analyze a year of posts against a goal, such as a 50/50 split between brand and B2B content, to plan next year’s editorial calendar.
• Surface risks, cultural sensitivities and timing issues for global audiences, like holidays or performance review week.
• Identify likely objections from legal or HR.
• Turn a 90-minute company meeting transcript into an executive summary and intranet post the same afternoon (an attendee favorite).

Her closing point: don’t measure the time AI saves. Measure what value it added and what you did with the time. And some things stay human, including organizational context, relationships and your voice.

The intent gap: the message you send isn’t the message they get

David Bowman, Product Director at Fresh Intranet, shared findings from the company’s Employee Attention Recession research, a February 2026 survey of 1,000 people in the U.S. and UK, split evenly between internal communications professionals and the employees who receive their content.

• 95% of employees trust AI to accurately summarize internal communications.
• 92% of internal communications professionals worry AI summaries distort their meaning or tone.
• 88% of employees have used AI to summarize content, and 58% do it regularly. AI summarization is now more common than reading in full.
• Only 33% of organizations have a formal way to monitor how their content is reinterpreted.

Fresh calls the distance between what you meant and what lands “the intent gap.” Bowman offered three levers:

• Strategic subtraction. Remove content and campaigns before adding more.
• Design content to survive mediation. Assume a long CEO memo will become a few bullet points, so be clear and put the actions near the top.
• Govern how content is interpreted. Employees increasingly “pull” answers by asking Copilot nuanced questions after a town hall, instead of waiting for the follow-up email.

His test is easy to run today. Ask your corporate AI a real, nuanced question the way an employee would. No Copilot? Type a full question into SharePoint’s enterprise search. “If search is struggling with that today, Copilot is definitely going to struggle with it tomorrow,” because both read the same content.

Earned media in the age of AI: why PR now drives LLM visibility

Tiffany Guarnaccia, CEO and founder of B2B tech PR agency Kite Hill PR, explained how AI is reshaping earned media, and how her team uses it to amplify every win.

The new media mix

Earned media now spans Substack, podcasts, YouTube, LinkedIn and even Reddit, which is emerging as a B2B channel. As newsrooms shrink, many journalists are taking their audiences to podcasts and newsletters, while those who remain are asked to produce more formats. Guarnaccia advised caution on LinkedIn as AI-generated “slop” grows, and said the fix is authenticity and a sharp red pen.

Earned media and AI search

Research Kite Hill has done with partner SUSO Digital shows earned media is among the top sources driving brand visibility in large language models. LLMs weigh relevance, credibility, clarity, consistency, third-party validation, domain authority, volume and recency. Trade publications can outweigh business media depending on your company. That gives communicators a new way to show executives the value of PR.

How to write a press release for AI search

Kite Hill now writes some releases specifically for answer engine optimization (AEO), and sends those to the wire, not to journalists:

• Use clear headlines and bullet points for readability.
• Frame the content around the questions people ask. Think in prompts.
• Add an FAQ section at the end, such as “Why does this product matter?” and “Who is it for?”

Amplifying every earned media win

When a story lands, Kite Hill runs an AI workflow, trained in the company’s and the executive’s voice, to adapt it for the blog, website, sales team and newsletter. It often adds a small paid LinkedIn boost of around $2,000 targeted to ideal customer profiles. The team uses Google Gemini and an agency platform called Suits, and built a custom “AI red flag detector” that checks sources and data and flags copy that sounds like AI. Guarnaccia compares it to a QA department for communications. Her bottom line: “AI is the enabler, and it does not replace the need for human creativity.”

Frequently asked questions

What is the intent gap in internal communications?

It’s the difference between the message a communicator sends and what employees understand after an AI tool summarizes or reinterprets it. Fresh Intranet’s research found 95% of employees trust AI summaries while 92% of communicators worry those summaries distort meaning.

When should communicators use AI and when should they avoid it?

Use AI to augment your strengths and supplement repetitive, low-value tasks. Keep it away from high-risk work involving reputation, security or regulated data, and apply human judgment wherever organizational context matters.

How does earned media affect AI search visibility?

Earned media coverage is one of the top sources large language models cite, so PR wins can improve how often a brand appears in AI-generated answers. Releases structured for AI readability, with bullets and an FAQ, can help further.

Upcoming AI events for communicators

Keep building your AI skills at the AI for Government Communications Virtual Conference (October 14-15, 2026), AI Strategies for Corporate Communications (October 27-29), the AI for Communications Virtual Masterclass (December 10) and the Practical AI for Communications Virtual Conference (January 20-21, 2027).