Privacy and Data Protection Hits the Big Time at VivaTech 2018
The announcement that Mark Zuckerberg would speak at this year’s VivaTechnology Conference was met with mixed reactions–and even a little smirking. Zuckerberg’s appearance, of course, would be the day before the EU’s General Data Protection Regulation (GDPR) came into effect.
In the last six months, consumers and businesses alike have come to terms with just how much of their personal information has been swallowed up by companies and business leaders like Zuckerberg suddenly find themselves under the microscope. Google has even dropped it’s “Don’t be Evil” mantra.
Although it’s impossible to know the full impact of GDPR right now, many of the points it addresses were already on display at VivaTech–especially in the world of AI and Voice Assistants.
Privacy, Security & AI
With the buzz about GDPR reverberating throughout the conference, it’s unsurprising that a common theme at VivaTech 2018 was the issue of trust and privacy as AI comes of age. How do we train and program AI so that it behaves ethically and protects users? How does machine learning function when rampant data collection is becoming increasingly frowned upon?
Once seen as arcane, questions revolving around the intersection of privacy/security and AI were front and center at VivaTech. Indeed, there was almost a tangible whisper from crowds as expert after expert weighed in on topics ranging from context aware assistants that serve appropriate messages at the right time to AI that can help doctors improve surgeries and recovery times.
At the center of innovation in AI is the fundamental importance of privacy and data ownership. Our CEO Rand Hindi debated many of these issues on a panel with the likes of IBM and Weborama, highlighting the shifting emphasis for companies in AI and beyond. “The legal issue is easy to solve. It’s establishing a mindset of privacy that’s the real challenge,” Rand argued. As with any foundational cultural movement, the GDPR has forced AI companies to think more long-term about how their products and services adapt to a world where compliance is required–instead of overlooked.
Voice Is the New Standard Human-Machine-Interface
Within the broader world of AI, voice assistants are quickly becoming the dominant paradigm for humans to interact with technology–for consumers and businesses alike. A recent ComScore report noted that 50% of all internet searches will use voice by 2020.
At VivaTech, dozens of companies openly grappled with how the spread of voice Human-Machine-Interfaces (HMI) will change business operations. Massive voice assistant adoption fundamentally changes the way traditional business is done for many companies. At VivaTech it was clear that this means universal practices like marketing/customer success, user experience and design are impacted. The shear number of voice devices showcases the scale of growth. Google and Amazon each sold nearly 10 million smart speakers in Q4 2017 alone.
We built Snips to make technology disappear by adding voice to every connected device. What does the standardization of voice assistants mean as the future we’ve been working towards becomes a reality?
Post-GDPR Voice and AI in Europe
Although GDPR has sent ripples across the Atlantic, the brunt of its impact has been felt in Europe. Until recently, the AI industry adhered to a centralized paradigm where user data was stored and processed in the cloud by large (typically American) tech companies. The rejection of this system proved an instrumental motivator to take back data sovereignty that spawned GDPR in the first place.
Yet the proliferation of AI has precipitated questions of data ownership and individual privacy that, quite simply, didn’t exist before. Voice HMIs actually raise the stakes for sovereignty and compliance. With nearly 2.5 Billion voice devices expected to be in use by 2021, the urgent need to secure voice data has never been higher. Users voices are biometric data, making it even more crucial for companies to understand what happens with voice queries and data after they’ve interacted with a device.
Perhaps this means that users’ voice data will one day contribute to their Universal Digital Profile where they can control directly which pieces of their digital lives companies can access. Or maybe voice will need to be legally stored on-device (our approach at Snips) to protect users from unwanted breaches or intrusive advertising. Whatever the end result, it was abundantly clear at VivaTech that this is a beginning of the fight to protect users’ data. Not an end.