Jabra sees AI changing how people interact with workplace technology. But the company does not see itself competing with companies building large language models (LLMs) or AI agents. Speaking to The Times of India at InfoComm 2026, Calum MacDougall, President of Jabra’s Enterprise Division, explained where the company believes it fits as voice becomes a more important interface for enterprise AI. MacDougall joined Jabra-parent GN in 2015 and took over leadership of Jabra’s Enterprise Division in 2023, following senior marketing and commercial roles at Sony Mobile Communications. His focus today includes enterprise collaboration, Voice AI, hybrid work and how professional audio and video are evolving for AI-enabled workplaces.“We are not the right people to do LLMs. We are not the right people to create agents. I think that’s up to other companies and people,” MacDougall said. Instead, Jabra wants to focus on making the interaction between people and AI accurate, seamless and secure, from the microphone and headset to the software managing those devices.While this puts Jabra at the hardware end of a rapidly expanding AI ecosystem, MacDougall argues that this position remains important. “So I think at the very end of the value chain, but we’re still in the value chain.” He also discussed multi-agent workflows, enterprise security, deepfake detection and privacy concerns around AI-powered workplace audio. Read the full interview here:Q: All your competitors are now telling the same “AI-ready audio” story. What can Jabra actually do differently, beyond a better microphone array?A: “I think maybe the starting point is: what do we think the AI future will be, and then how do we prepare for it to help our customers and partners? So I think we believe agentic workflows will change the way everybody works, and that voice will be a key part of that.You can talk 3 times faster than you can type. It’s more natural to talk. Particularly if you’re talking to an agent or an AI agent, the more context you give it, the more effective it is.So, if you try making a PowerPoint prompt by typing and then try making a PowerPoint prompt using your voice, the voice version is better. So, if we believe in that voice world, the 1st part actually is the microphone.You need an accurate microphone. So if we look at the evolved 3 range that we mentioned before, and if we look at the accuracy of voice transcription, our products are 98% accurate if we’re looking at copilots.If you’re looking at consumer products, accuracy is lower. So it’s an important starting point. That’s where we stand today. In the future, we want to add a few other parts. We talk about being accurate, seamless, and secure.By seamless, we mean being able to easily talk to whichever agent you want. In the future, we believe you won’t want to talk to just one agent; you’ll have 2 or 3. So this is what we’re currently working on in development now: creating a really simple way that you can be in a meeting, but then potentially also talk to your agent as well. Not in the market now, but something we’re developing in the future.The other part is data security. The company wants to keep their data secure, so we help enterprises make sure employees use only the right agents and access only the right information. That’s one part of security. The other thing we’re working on for the future is how we can combat deepfakes. For example, you take a phone call, and someone asks you to transfer money, but it’s an AI fake. We’re looking at technology we can integrate into our products that can identify whether it’s a real person.Today, much of the story is about microphone accuracy; in the future, we can see a lot of fresh experiences we could bring.”Q: If voice becomes the default interface for enterprise AI, the real value shifts to whoever owns the AI layer, not whoever makes the mic. Doesn’t that make Jabra’s hardware a commodity in its own story — and where does that leave you in five years?A: “I think, first, hardware will still matter to deliver the right experience. And also, we talked about security; trust will matter- the ability to be able to trust the brand on the edge. But as we said, I think there are still opportunities to deliver value, not just at the hardware layer, but at the software layer as well. For example, imagine you want to talk to multiple agents; wouldn’t it be nice and easy if you could say, “Hey, Jabra, can I talk to ex-agent or Y agent?” So I think we can deliver some value there at the software layer.We are not the right people to do LLMs. We are not the right people to create agents. I think that’s up to other companies and people. Our job is to really be the secure, accurate, seamless microphone on the edge because you can spend as much money as you like on data centres and LLMs. But if the agent doesn’t hear you, and you can’t easily access the agent in a simple, secure way, people won’t use it. So I think at the very end of the value chain, but we’re still in the value chain.We have our own application software called Jabra Plus. You can use it as an individual to optimise your headset settings. And as an enterprise, you can use it to manage your headset portfolio. So imagine I’m a company and I have 5000 headsets out there. I may want to update the firmware or manage the security. I can manage that through Jabra Plus. In the future, we believe Jabra Plus has a role in this AI world. Enterprises can monitor and manage which AI agents people use. And for me as an individual, perhaps there are 3 or 4 agents I would like to use. If I could integrate them into Jabra Plus, then I would have this seamless experience that people talk about.”Q: Always-on, AI-optimized microphones in the workplace raise an obvious question: who has access to that audio, how long is it kept, and what stops an employer from turning “convenience” into ambient surveillance? What’s Jabra’s actual technical position — not the reassurance line?A: “We can talk a lot about the technology, but there is a cultural and privacy element here. The notion of being always recorded, and what happens to data, legitimately raises concerns for people. Some of the things we are discussing with customers are that, at least on the devices we have, there’s some form of notification so people know when they’re being recorded.For example, a red light comes on so you know that you’re being recorded. One use case we’re working on is in the medical sector, where AI can support doctor-patient consultations and note-taking. For example, in a consultation, I’m the doctor; I don’t have to take notes anymore. I can talk to you. Then I will take the notes and maybe do the referral. But you would like to know that it’s being recorded in some way. So we have a ‘Speak’ device with a red light on it, and we’re looking at portable microphones in a similar way.So I think the notion of data and privacy is important for individual company policies, software policies. I think we can help on the edge by showing that you’re being recorded and included in the value chain that way.”