How EAs Are Using AI Every Day

There’s always a shiny new AI tool. We’re constantly hearing about what’s possible.

But if you’ve ever thought, “Okay… but how do I actually use this in a way that’s useful day-to-day?” — I hear you.

So I reached out to some brilliant EAs and leaders in my network to see exactly how they’re using AI in their daily work. The response? Practical, honest, and genuinely useful.

No hype. Just real examples from real assistants.

Keep reading to see how EAs are actually using AI every day.

Hillary Robertson 

As a Fractional EA, I do a lot of networking and meeting in coffee shops. I was using a pretty good prompt a few times to find an independent location midway between me and my guest, and decided it should be an Agent and do more than I originally asked the prompt to.

The Coffee Shop Concierge Agent delivers the top 5 locations (website, location, brief ‘vibe report’) midway between you and your guest that will be open at the time of your meeting. It features highly rated independent coffee shops that are owned by women / BIPOC / LGBTQIA2S+ because these are who I want my dollars to support. 

The top location is closest to the equitable midpoint location. The bonus feature of this Agent is that if you say one of the attendees is a CEO/executive/founder, it will select options with a quieter atmosphere and better parking options, perfect for busy executives.

Candice Burningham 

I’ve been doing a lot of venue sourcing for upcoming conferences and events. I use AI to:

• Compare ROI across different locations
• Identify potential risks (travel, cost blowouts, timing)
• Pull insights from reviews across trusted platforms
• Suggest optimisations based on the brief

I continue refining my prompts and questions until I get to a clear output matrix that helps me quickly assess whether a venue is worth investigating further. 

Fiona Young 

Fiona is the Founder of Carve and she told us about some of the ways EAs who have taken her training are using AI daily. 

Never miss a task buried in an email or Slack again

Problem: So. Much. Noise.

Solution: Build a Power Automate trigger that detects the phrase "Amy can you..." and automatically adds a task to task list with context (owner, due date, link). 

Quickly scout out details buried in docs

Problem: Wasting hours each week searching across 47+ team documents to answer recurring questions about travel policies, vendor contacts, workflows and more.

Solution: Build an AI assistant that indexes the most referenced files and delivers cited answers in seconds. Eventually build a Slack integration.

Equip PMs to self-serve

Problem: Due to restructuring, some project work is losing EA support.. leaving project managers without admin or planning help

Solution: Build AI assistants for repeatable project-related EA workflows (starting with an Event Planner Task List) so that PMs can self-serve from a few key inputs (date, budget, attendees, etc). Eventually bolt on a Power Automate flow to auto-populate tasks in PM tool.

Track industry trends to support M&A

Problem: Leaders of a group of 150+ companies need rapid insight into industry trends buried in quarterly earnings calls across competitors, customers and vendors. Manually reviewing + summarizing these calls is time consuming + inconsistent.

Solution: Build an AI assistant to extract emerging trends and who said what from call transcripts. Once polished, launch as an internal Enterprise Custom GPT so other divisions can reuse it ➡️ a standardized way to capture market intelligence across the org.

Yvette Lamb (née Pearson) 

• I've taken a technical specification document (for salesforce migration) and asked to summarize for leadership sign off, highlighting the key trade offs that the MVP won't cover in favour of fast deployment. 
• I've created a Gemini Gem to review customer contracts. I uploaded the contracts and then I can go back and ask it questions, such as "when should I start renewal discussions?"
• I uploaded about a dozen conversation transcripts about how teams use our sales tech stack and asked it to summarize the system requirements for the salesforce migration. 

Jodie Mears 

First thing I'd say is don't hide it! I tell my exec and colleagues where AI sits in my workflow and what it adds. That's opened more doors than the tools themselves. Conversations about funding for training, trying new things, and helping other EAs get started. 

The more open you are about it, the easier it is for your organization to see you as someone who leads on this—not just someone secretly dabbling with it. When AI use is visible, it becomes a career conversation—not just a productivity one.

Start with what you already have. I work in Microsoft 365 every day—so it made sense to go deeper there before looking anywhere else. Most people skip past what's already in their toolkit. Excel, PPT, Power BI, Planner, Teams, all with Copilot built in now.

On compliance—this is always the bone of contention in conversations on AI use in corporate settings, and yes, it matters and should never be an afterthought. 

A full enterprise MS Copilot licence has it built in from day one. Set up for your organisation, works within your access levels, searches only within your own environment. You're not pulling data from outside or going around your security policies—you're working within them. That's why I'd always point EAs to the tools their organisation already owns first.

One automation I've set up that I genuinely couldn't go back from—I scheduled a daily Copilot prompt to land in my inbox every morning at 8am covering both mine and my exec's to-dos. It's like a start of the day write-up we used to pull together manually back in the day, and honestly, it used to take me ages- now it just arrives. The prompt asks for an EA-style summary, today's emails only, split into three sections. Actions and decisions needed, important awareness, and FYI / low priority. 

It cuts the so-called "noise" of the inbox, flags what needs attention and ends with a short takeaway. Totally skimmable in a few minutes. I call it my "view from the bridge", and it sets up the whole day before I've opened a single email thread.

Meetings are where I've probably felt the biggest shift though. Before, I'd be in a meeting trying to listen properly, take notes, keep track of who said they'd do what, and mentally flag the things I needed to follow up on afterwards. Something always needed clarity afterwards, now I go into every exec and team meeting with MS Planner already open. As actions come up, I log them in real time, owner assigned by me, due date set, right there in the meeting, and also send the action owners reminders. 

When the meeting ends, Copilot pulls together a summary of what was discussed, what was decided and what needs to happen next. I review it, add anything that needs context, upload it to where we store meeting notes, and it goes out in minutes. 

The loop actually closes, and the people do the things they said they'd do because there's a record and a reminder automated. That sounds simple, but it changes how an operation runs, and that's exactly where an EA should be making a difference.

I also use Copilot in Excel and Power BI to give my exec a live KPI view. One place that answers the question, 'Are we on track?' and 'Where are the risks?' Less time building slides, more time talking about what the numbers actually mean. For my coaching, training and speaking work I use a different tech stack—different outputs, different audiences, different tools for the job.

But in the EA role, for me—I keep coming back to the same thing. My value lies in the judgement and context I alone have. AI gets me to the starting point faster. I take it the rest of the way. 

Make it visible, tie it to decisions and follow through. Start with what's already sitting in Microsoft 365 (or your operating system you use) and treat it as a skill you're building, not a shortcut you're hiding. That's where it all gets interesting. 

Sarah Howson Marianne Whitlock 

As a small but mighty team, anything that helps us work smarter, create efficiencies and free up time for higher-value thinking has a significant impact. AI has become part of our day-to-day rhythm, supporting everything from recording meetings and capturing actions, to working through voice notes when we’re on the move. It allows us to stay agile, responsive and focused on what really matters, our community, our clients and being able to deliver promptly.

In practical terms, we use AI to help shape tricky emails, improve workflows and sense-check messaging when we need a fresh perspective, sometimes on a Friday afternoon after a busy week when our brains can be a little numb. That said, it always comes back to the quality of the prompt and, more importantly, the human touch that follows. We would never send anything out without reviewing, refining and ensuring it reflects the right tone, accuracy and intent. For us, AI is a starting point, a tool for inspiration and structure, but the final output is always carefully crafted so it feels authentic and aligned to our voices.

Ultimately, for us, it’s about balance. We will continue to explore how AI can drive efficiency in our businesses and elevate how we work, but never at the expense of authenticity or relationships. The human element remains at the heart of everything we do, and that is where the real value lies. AI can’t hug like we can!

Angel Ahlstedt

• I've automated daily and weekly task summary emails for different projects/groups that pull from my Notion database. I used Gemini, ChatGPT and Claude to help write code for this automation.
• I've built dashboards in Notion using AI as a thought partner first, testing it's assumptions by pushing back and having it challenge its own logic and the efficacy of its suggestions. Once I'm satisfied with the output, I use it as a knowledgeable companion to assist with the build itself. 
• I've created Google Gems that analyze transcriptions and train of thought notes to produce summaries and action items. These Gems have context for names of staff, roles, names of clients, etc.

Mia Olsson

Every month I need to split our PDF Amex statements into ~ 80 individual files - something Maxwell (my ChatGPT) does for me with a single prompt, saving the individual statement by the card holder's name in a zip file for me to download. 

Saves me a ton of time.

Jacqueline Nelson

At work: I've used Copilot to help me create an automation in MS Power Automate that sends me my daily deliverables that it pulls from a list I created in MS List.

Personal: I've created an agent with Zapier to categorize emails that come into my Gmail account and another to help me respond to people who reach out me. This second agent looks up the sender’s online profile, sends me a summary of who the person is and then drafts a response according to the nature of the sender’s email and saves it for me to review and send later. 

Alicia Fairclough 

Before I began writing a job description, I turned to ChatGPT. I gave it context about the business—where the revenue comes from, what support I already had in place, and how I was feeling about my workload and where things felt stretched. Then I asked it to help me figure out where hiring would have the most impact.

From there, we worked through the job description together. It took a vague feeling of "I need help" and turned it into a concrete, well-reasoned role.

Rhonda's Closet 

A more personal example is my Custom GPT Rhonda's Closet. This is what I use to quickly and easily create Vinted listings. 

I dress my mannequin Rhonda in the item of clothing I want to sell and I take a picture. I upload the photo along with a picture of the tags to 'Rhonda's Closet'. 

Rhonda's Closet quickly provides an accurate Vinted description in the exact format I want complete with title, materials, size and other important details. It has saved me loads of time! 

The Executive Assistant Interview Companion 

I created a CustomGPT based on The EA Interview Blueprint, a resource I created for EAs to help them prepare for interviews. 

You can use it here! 

The through-line in all of these? AI rewards specificity. Bring it the specifics and it pulls its weight. Keep it vague and you'll be editing garbage.

The best EAs I know aren't using AI to do less. They're using it to finally have enough bandwidth to do everything they've always been capable of.

We love hearing how EAs are putting AI to work in the real world. Leave a comment telling us how you're using it day-to-day.