#312: Focus on Sustainability in Addressing Gen AI Ethics and Risks: Prakash Arunkundrum, Logitech

Great, Dr. Very nice to see you again. It's been a real pleasure to be talking to you. So I am Prakash Arunkundrum, President of Logitech for Business. Logitech for Business is a division of Logitech that makes products for customers, enterprises large and small, from video cameras to mice, keyboards, headsets, webcams, and everything in between, including accessories and peripherals. So all of those things are part of our Logitech for Business. I've been at Logitech now for 10 years, in various roles. Until recently I was the Chief Operating Officer for Logitech, also ran operations, product development, manufacturing, supply chain, and also led our sustainability efforts, driving carbon reduction across our entire value chain. So that's a little bit about me, but super excited to be here.

Yeah, I mean this is by far the most transformative time that we as a species are going through. There's been a healthy level of questions, curiosity, and innovation happening within Logitech. Some of our product teams, like video conferencing equipment, already deploy machine learning, and now with advanced models they're doing more. The product teams were excited to get on GitHub and do more with software. Our marketers started exploring how to make better icons, video, and avatars. We made policies around product images and avoiding areas around people and personalities. Our employees wanted to use the latest AI tools. I personally use at least three or four different tools every day inside our network. We created that to protect our IP but also use the best of AI. On average, our employees are super excited. As a tech company, this is the most transformative thing happening. Internally, we're super excited, seeing broad adoption in operations, marketing, software engineering, product teams thinking about AI in the context of products. It's a fun time to be in the midst of this.

Yeah, I'll start with leadership and then the people in the team. Leadership had very little resistance to AI, stemming from our history as a tech company. We understand we have to innovate to keep up. The biggest topic for us was how to protect privacy and proprietary content. That was one place we spent time on. We've been using AI in video conferencing well before it was a buzzword. We had RightSight and RightSound announced before 2022. It's been part of our products. We've had a machine learning team looking at computer vision for at least eight years. As models got more capable, it opened new possibilities. We didn't get a lot of job loss discussions. The bigger questions were how to prevent bias, ensure the right data, and understand the input data. Logitech plays in that world of input – audio, video, and input from mice and keyboards. So we are aware of our role in the ecosystem, which brought awareness around Swiss neutrality, privacy, and bias aspects. That was the biggest area of focus. We've not really received too much internal resistance. There's been more want to try the latest. We set up an AI governance board that approves models for internal use, have responsible AI policies, and signed up for the European Responsible AI Act.

Yeah, I'll start with the most basic example. When a video camera recognizes me to represent me in the best light, the models need to understand lighting, tone, color, complexion. Same for audio. We spent considerable time making sure the data models for RightSight and RightSound are based on as wide and diverse a set of parameters as possible – differences in tones, accents, pitches, male/female. That's top of mind. Also, having access to data – who gets access – is clearly managed by our legal team and Chief Information Security Officer who oversees privacy and security. We treat AI as an additional gate in our security and privacy processes.

Actually, in our recent annual investor day, we shared our AI strategy as part of our mission. It goes back to a central theme: people want more precision, more intent, and more smarts. As a company, we make simple, smart, and sustainable products. The domains where we apply this are audio, video, and input. In the ecosystem, it's our job to ensure that the data going to cloud models has that last millimeter fidelity. We defined that as our core place in the ecosystem. We're not going to become a language model company. Logitech makes products for work and play. Our mission is to extend human potential in work and play, making the human more capable. That's the other side – making the human more important, making audio, video, and inputs more important. That's where we reemphasized our core values.

Yeah, this is the most important thing – trust and credibility with customers using AI-driven features. We signed up for the Responsible AI Act from the European Union for transparency and accountability. Second, regarding private data, even before AI we were mindful about what private information we store. We don't store personal information as a policy, so signing up for the Act was easy. Third, working with ecosystem partners like Microsoft, Zoom, Google to define roles and data protection. Enterprise customers rely on cloud providers; we ensure our devices provide additional functionality in service of those platforms.

This is huge. As a company, we've focused on sustainable practices for a long time. We exist in the context of an ecosystem. First, we clarify our own position – what bias looks like and what we stand for, which is shown in the kind of people we bring and our equality rules. It starts from sustainable practices, which lead to fairness and equity. Second, we have safeguards: a creator and a reviewer role for models and data. This is natural to us as a hardware company with quality criteria. That's really the two core things: sustainable practices and being mindful of who accesses data. There's still more to do as a society to reduce bias. We're not perfect, but acknowledging we have work to do is important.

Yeah, it's useful to talk about Logitech's products and how they intersect AI. Starting with video bars: we have Logitech Sight, a bar in the front of the room connected to center-of-room cameras. We can connect two center cameras – six cameras and twelve microphones – using machine learning and cinematic framing to create equitable meetings so remote participants feel present. That got Time's Invention of the Year award. Another example is a two-way AI headset that cancels noise on both sides by filtering only human voice frequencies. Third, mice and keyboards that let you quickly launch your favorite AI tool in context, like paraphrasing documents. Recently, with our Streamlabs capability on gaming, we use agentic AI as a sidekick coach for gamers. So we're augmenting human workflows, making technology disappear and just work like magic.

Thank you, Dr. Gleb, and thank you to the audience for checking out another episode of the Wise Decisionmaker Show. Please make sure to subscribe wherever you checked out the show and leave a review. It helps other people discover the show and helps us improve. If you have any questions, email me at [email protected].