This time last year we had a dilemma on our hands. AI was taking off and we wanted to give our team the freedom to experiment with it.
But everything you enter into the free version of ChatGPT (for example), is used to train the AI and can end up in the public domain.
Faced with this predicament, lots of companies banned AI, or created complicated rules for their staff to follow – to protect their privacy.
For all the benefits without the data privacy risks, we decided to make our own private environment (we called it Definition AI – and we’re building them for clients too).
And here’s how our specialist teams are using it
“I use our AI to help with concept development and brainstorms. We’ll often put a client brief into our platform, tell it what we’re looking for and ask for some ideas. Then we go through and either polish those ideas or ask it for more.
“Sure, not all of it will be brilliant. But it’s enough to spark some creativity and we can develop a concept from there.
“A private environment means we can add the actual brief from a client, including all the nitty gritty bits that aren’t for public eyes. So, our AI now has accurate context to base its responses on, which means better output.”
Luke Budka – AI
“I use it for creating training content. AI makes it easy to come up with practical exercises and examples for our training sessions.
“And we keep it private because training is a product we sell. We work hard on the contents and don’t want to give it away for free.”
“I use Definition AI to create lists of potential article titles. It’s great for getting past the copywriter’s nightmare: the blank page. And then when I’ve got something I like I can use it to refine those ideas, too.
“Obviously you don’t want the public seeing anything before the client approves it, so it has to be private.”
“AI’s great for automated reporting. For example, using generative AI to analyse data from Google Analytics or Search Console.
“Doing this in a private environment means we can avoid any sensitive performance related metrics being shared publicly.”
Caitlin Singh – Business services PR
“I use Definition AI for predictive analysis like spotting industry trends. So it needs to be private because we wouldn’t put a client’s internal sales data, existing business trends and consumer behaviours where others could use or see them.”
Olivia Prole – Business services PR
“AI’s my assistant. I put in meeting notes and Definition AI gives me a summary and a prioritised to-do list. Obviously I’ll go on to tweak it myself, but it’s a secure space to speed up my thought process.”
“I use Definition AI to help write social media copy in the right tone of voice. I write the content, then I train the AI with client documents that are already pitch perfect, and then I can go through and edit from there.
“News changes quickly, so there’s no way we’re making client social content public before sign off. That’s why our private environment is indispensable for me.”
Jasmin Martin – Media relations
“Definition AI is helpful when we’re working on pitches for journalists. We can input a press release, and it helps guide our pitch based on the key details. Helping us analyse the story from different perspectives can spark ideas of different ways to get that coverage.
“Loads of our press releases have sensitive information, so using our locked down environment guarantees that no pre-pitching leaks will happen.”
Mitchell Foot – Creative marketing
“Definition AI acts as my sidekick. It helps me analyse loads of data for trends, key points and insights, as well as providing a different perspective on campaigns and sparking ideas.
“Privacy is important, given the sensitive nature of the data being processed. Our PE gives me a sense of reassurance.”
Get involved
We don’t just build private environments for our business. We do it for clients too. It’s a secure space to do all the above and more – and it’ll cost a lot less than having dozens of OpenAI licences.
Just give us a shout and we can talk it through.
