When we established the Talacom AI Working Group earlier this year we had a simple but ambitious goal. With AI’s continued rapid growth across all industry segments, we needed to broaden our understanding of this technology and explore ways that we could responsibly harness it to deliver greater value for our clients.
AI’s transformative impact cannot be ignored; it’s revolutionising the workplace, and people are quickly becoming accustomed to using it. Just like email, social media, and virtual meetings before it, AI will become a foundational business tool within the next few years. To stay ahead, we must invest not only in the right tools, but also in upskilling our team to use them confidently and responsibly.
Though a few team members had experimented with tools like ChatGPT and Gemini, the newly formed working group was far from a seasoned tech taskforce. It was a cross-functional team of curious minds, eager to explore how AI could enhance our work without compromising our values.
Cautious curiosity
“Move fast and break things” might work for Silicon Valley, but it can be a recipe for disaster in communications. Trust is everything; it is the foundation of reputations and relationships, and trust accumulated over years can be lost in an instant due to a single foolish decision.
The working group’s job was to help us move from curiosity to clarity, from simple experimentation to strategic execution. Ultimately, this workstream needed to equip us with the tools and framework to embrace AI in a way that would be responsible, transparent, and client focused.
We asked ourselves: How can we harness AI without losing the human insight, empathy and sectoral expertise that define our agency? How can we securely and responsibly introduce AI tools and protect the privacy of clients and our team? And how can we continue to broaden our understanding of AI as technology and applications evolve?
In addition to seeking external expert insight, we also surveyed the team to get a handle on their perspectives. As an employee-owned business navigating how AI is reshaping our work, we knew that we had to bring the whole team with us on this journey.
Our investigations and subsequent deliberations allowed us to coalesce around three core principles, and we have now publicly committed to:
- Being open about how we use AI tools in our workflow
- Using AI as an expert assistant to support our team, not replace it
- Nurturing our understanding of AI and its ethical usage
You can read more about these principles and how we intend to uphold them by reading the Talacom AI Vision Statement.
Turning principles into practice
With our principles in place, the next challenge was translating them into action. A set of AI Usage Guidelines translated this collective vision into practical, day-to-day behaviours, while also outlining our procedures for vetting AI tools, and reinforcing the importance of training our team on how to spot bias and misinformation.
We’re under no illusion; this is a living document that will continue to evolve to keep pace with technology and a growing appreciation of ethical considerations around AI. We knew waiting for perfection would stall progress, so we focused on building a solid foundation we could refine over time.
Without a set of internal policies, we couldn’t look our clients in the eye and say that we were acting responsibly and with strategic forethought.
Investing in enterprise grade technology
Ok, hands up, at the first meeting of the working group we couldn’t resist talking about the concept of “TalacomGPT” and building our own tech stack using some of the open-source models.
While this would have made for some great agency marketing content, going down this road would have been a huge mistake. Even if we had the in-house technical expertise and budget to pull it off, we couldn’t define how a bespoke solution could offer significantly more than something off the shelf, plus it would rapidly become outdated.
After evaluating various options, we soon concluded that investing in Copilot 365 licences made the most sense given its integration into the Office 365 applications that we all use day to day, and its ability to digest and ground responses in our existing data was a huge benefit.
Most importantly, Copilot 365’s enterprise-grade security, encryption of all data, and guarantee that confidential information is not used for training AI models gave us confidence that it would be safe for handling sensitive client and agency matters.
We’re now just a couple of weeks into the team-wide rollout, but Copilot is already heavily embedded in our workflow, helping us work smarter and faster, but never at the expense of quality or integrity.
I’m sure it won’t be surprising to learn that this blog was drafted with assistance from AI. Given probably only 20% of what it suggested has made the cut, I blame the human who briefed it!
Next steps
We know that we’re still in the relatively early stages of our AI journey. As we look ahead, we’ll continue to innovate with purpose. We’ll keep asking hard questions. We’ll keep refining our policies. And we’ll keep putting trust, responsibility and opportunity at the heart of everything we do. Because for us, AI isn’t a convenient buzzword to throw around in a pitch meeting, it’s an opportunity to deliver better outcomes for our clients, and to do so with integrity.
PR agencies rebranding and putting AI at the centre of their proposition risk losing sight of their true purpose and being accused of boosterism. Our role is to forge and maintain relationships, and people will always be central to this mission. “Human first and human last” is a maxim that we stand by – we’re not outsourcing our emotional intelligence to a machine.
We hope you’ll join us on this journey and we’d love to discuss shared learnings with clients, partners and peers as we shape the future of responsible AI in communications.
Andrew Baud, CEO Talacom