The comms professional's guide to surviving the AI revolution
Axios's Eleanor Hawkins on AI tools, disclosure dilemmas, and why relationships matter more than ever
🇹🇷 Istanbul’s annual contemporary art fair is on this week, and the conversations are mercifully free of AI speculation. It’s a reminder that AI chat in the corporate world is still mostly an echo chamber.
👩🏻💻 This disconnect sits at the heart of our conversation with Eleanor Hawkins, Axios writer and communications strategist, who made the rare move from comms to journalism. She writes the outlet’s weekly newsletter for communicators, and has a unique view of how well the AI rhetoric matches reality, as both an AI user and industry observer.
Why it matters (Borrowing Axios’s Smart Brevity style!): While 80% of corporate comms work can be augmented or automated with AI, success increasingly depends on tapping into the right audiences and relationships. The risk for CEOs and communicators lies when narratives to different audiences don’t match up.
EW: Do you think the latest AI tools, powered by large language models, are truly revolutionary for communications?
EH: I do think it's totally revolutionary. There's a lot of talk about how AI will replace communicators, but I don't know that I see that. What I think it will do is totally change the way that they work.
A BCG report found that 80% of the work of corporate affairs — comms typically fits under corporate affairs — can be either augmented or automated using AI. That's an incredibly scary number for most comms leaders, because they think, “Oh, AI can do 80% of our work. That means that we don't need as much headcount, as much budget.”
But really, what it comes down to is a lot of time savings and using AI to quickly generate briefing binders, for example. All of the information that you put into it, all of the prompts, all of the historical knowledge, that is coming from a human.
Right now, it is just generative AI. I think what will really save a lot of time and money for teams is once agentic AI is more mainstream.
EW: What’s best practice AI use in comms teams?
EH: There's no best practice yet. Right now, a lot of folks are just playing around. But the companies that are really making headways — the order is coming from the CEO.
Cognizant is a great example. They had a vibe coding event where their comms team was able to create their own tools to help solve a lot of their problems, like measuring reputation, measuring corporate narrative, and being able to tie that back to share price.
The Instacart team uses AI to handle news clips. If you've ever worked in communication, that is usually the first job you are given, and it takes a long time. The benefit is, it allows you to understand what's happening in the news, what your competitors are doing and how the media is covering your company, your industry.
Instacart has created an AI tool that does that for you. Then it is incumbent upon that junior member to pick and choose what is strategically the most important story or narrative. It allows that person to connect the dots, as opposed to being bogged down by the tactical task.
EW: What will AI’s greatest value add be for communications teams?
EH: The most value-add is going to be the time saving.
Another thing that I think AI will help with is targeting audiences. There's a lot of talk in PR around like the ‘R’, the relationship piece. It's going to become even more important in the age of AI.
The public piece is really important too. Are you trying to reach the public with your message? Chances are not, so having these AI tools that can help you identify key audiences and message to them. I'm seeing a lot of that: AI helping you drill down on who matters and why and who moves the needle where.
EW: How is Axios using AI?
EH: Our CEO, Jim VandeHei, and our leadership team here at Axios are very bullish on AI. We candidly have a partnership with OpenAI, the caveat being that we don't use AI to write our stories. Every story you read on Axios is written by a human.
Our stance comes down to the fact that we want readers to trust and know that what they're reading and absorbing is coming from one of our experts in the newsroom.
We like leaning in on expertise, but believe that AI can help support the work that we do. For example, we have several company-wide GPTs that can help us come up with quick captions and Axios style. We have a very specific Smart Brevity style format, so it can help format us. I use it a lot for SEO, URLs.
EW: How else do you use AI?
EH: Our marketing team sends out a lot of copy that is signed by some of our experts. For example, they've created an Eleanor GPT. They take some of the emails that I've written in the past, or some of my reporting, and they can use it to create my voice.
I use [AI] a lot for speed editing and copy editing before I send it to our human copy editor. It saves her a lot of time, and it makes me feel confident in the product that I'm delivering.
I will sometimes use it if I'm doing more strategic comms work. I will drop in a top-line messaging doc and say, “Poke holes in this. What am I missing? What are some questions key stakeholders might have?” Sometimes I'll do the same when I'm prepping for an interview: “I'm interviewing this person. Here are the topics. What am I missing? How might they respond? What should I be aware of?” The same way that comms teams use it to prep their execs, I use it to prep as well.
Also, a great tool that we use across our newsroom is Otter, our transcription tool.
EW: Yes! I am a huge Otter fan.
EH: We were at a big festival earlier this summer, and I had 14 or something interviews back to back. I had 30 minutes to go back to the Airbnb, change, and get ready for whatever was next. I had maybe done half of those interviews already. So I took the transcripts, put them in ChatGPT, and said, “Write this in Axios style” while I was getting ready.
I came back and saw the finished product. It was not good. It was never anything that I would repurpose or use in my own reporting. But it helped me find a common thread. It pulled together these conversations and came up with a big picture story, a big trend story, in a way that I didn't have time to sit down and think about at that moment.
EW: What are some risks around AI that you’re seeing communications professionals grapple with?
EH: The first thing that I've heard is that most companies don't have clear policies when it comes to AI use. But even if they have policies, maybe they're not clearly communicated across the organization.
Also, disclosures — when you use AI to write something. We disclose when we're using our iPhone, for example. Are you going to disclose when something is sent using AI? I don't know, and I don't know if there will be any sort of legal ramifications for that.
There's also a bigger story here around job displacement. Something I'm paying a lot of attention to right now is how CEOs are communicating about AI specifically in their earnings. The message they're putting out to shareholders is all about efficiency, cost savings, etc. What they're saying to employees doesn't really mirror that. I do think how they message AI now will matter in the future, and anytime American jobs are at stake, it is inevitably going to become a wedge issue. What you're saying now could really matter in the future.
EW: Beyond the day-to-day tools, what are the bigger strategic challenges AI creates for communications teams?
EH: When I think about AI and communications, there are a few different buckets that I put it in. There's the work, which we've talked about, how AI is disrupting the work of a communication professional. There's the risk and the reputational risk that comes with being an AI-first company.
Then there is the belief that every company wants to become an AI-first company. So how do you market AI? How do you get people on board? How do you ‘PR’ intelligence? All the AI labs are doing that, but even if you're not one, there is this urge to want to be an AI-first company. How do you sell that, whether it's selling it to your customers or to your employees or to shareholders? That's a challenge for a lot of communicators, especially since few are able to show the ROI.
At least right now, it's not a tangible thing, so it's hard for people to wrap their heads around. We're very much in a bubble here talking about it. The average American doesn't know what agentic AI is, so there are going to be some communication challenges there.
The other thing that I'm really thinking through is GEO [Generative Engine Optimization].
EW: Fascinating! In my comms role with Bek Ventures, I’m seeing founders think a lot about this. Traffic is falling as more people search for information on AI tools, but funnily enough, those AI tools are pulling from traditional journalistic sources too.
EH: One of the learnings from this Muck Rack report that I thought was so interesting is that a lot of these LLMs are also pulling from non-traditional news sources, like podcasts, for example. Gemini specifically is most likely to pull from podcasts that also operate on YouTube, so if you’re booking a CEO on a podcast, make sure you also have the podcast run on YouTube, so it’s more likely to show up in the LLM. Little pointers like that are so interesting.
Also recognizing that owned content remains important, and not just your own content, but content from other corporate sites. An example of that would be an interview that somebody did with McKinsey. McKinsey will be viewed as a well-established source. So while that's not an editorial interview, it is an outside conversation that will influence these LLMs.
EW: What superpowers will comms professionals need in the future?
EH: It’s going to be partly [about connecting the dots]. I think it’s also going to be about understanding the key relationships, who you need to engage with, who moves the needle, and how to reach them. Almost thinking of communicators as those liaisons for each community that matters.
Reading this week
📚 The Times writer James Marriott argues we’re entering a post-literate age; investor Sarah Drinkwater raised the same concern in our recent interview.
✒️ AI commentator Azeem Azhar told us last week he had bought his team fountain pens and notebooks. I shared my own fountain pen story, which got an overwhelming amount of beautiful responses! Should we be bullish on pens?
😬 AI workslop is destroying productivity.
🧠 ChatGPT isn’t making us stupider, but we should still be worried. Good piece from Celia Ford on Transformer.
💖 A final note: some of our readers have asked to know about how others are using AI to write in second languages. If you’re using AI to write in a language that is not your mother tongue, share your experience in the comments or via email.
This edition used AI for subject recommendations.
Loved this interview. 100% you need the original thought to come from a human who has competence and capability in comms to input and also to critique the output.
Once the language model has examples of the work you can produce, your tone, etc then I think it can be used to mimic aspects of your writing to save time and discuss more ideas.
Fantastic interview, thanks so much for this. In particular I found it very revealing to see the mixed messages by CEOs surrounding AI use, depending on the audience.