Securing the Future of Identity with AI

"With artificial intelligence and machine learning, everything is dynamic…"

“With artificial intelligence and machine learning, everything is dynamic. Take, for example, employee onboarding. Machines don’t onboard everyone the same way: they have automated detection like automated criticism, or dynamic decisions. This is why the directory piece needs to be very dynamic, very scalable, and very response-driven. That's the fundamental notion. We’re going from a model where you know everything about everyone, to a new model where you don't know anything about anyone. AI is critical to managing that change.”

Venkat Sathyamurthy

OneLogin Chief Product Officer

SEE FULL QUOTE

"I think the biggest end-goal for AI-driven authentication is that there's no password…"

“I think the biggest end-goal for AI-driven authentication is that there's no password at all, and that the AI is so good that it can predict or retrieve the identity of the user with 99.999% accuracy. Then, you can use it for full identification, authentication, and for anything else that you want to use it for. But it's going to be really hard to get there. And, will we get there in five years? Maybe.”

Rich Chetwynd

OneLogin Product Manager, Developer Experience

SEE FULL QUOTE

Just the managing of your IOT devices is by far the biggest challenge

Just the managing of your IOT devices is by far the biggest challenge

WATCH VIDEO

"If I'm making you do MFA every single time you login, that's a horrible experience..."

“If I'm making you do MFA every single time you login, that's a horrible experience. The only way I think that organizations are really going to start adopting it [MFA] more is if they can find ways to increase their risk coverage, but not require users to have to do anything all the time. By using technology like Vigilance AI to really understand what other variables we can consider to verify that you are who you say you are, without having to make you prove it, that’s really going to be the key to MFA. Adoption requires using things like adaptive authentication and Vigilance AI to really help with process.

Jeff Broberg

Sr. Director, Product Management

SEE FULL QUOTE

"Ultimately, what's powering risk engines is machine learning..."

“Ultimately, what's powering risk engines is machine learning. It’s absolutely important for balancing trust and experience in the identity industry. It's still immature, though. There is a lot of science and experimentation that goes into artificial intelligence (AI) and machine learning (ML) work. AI/ML is not something that you really define with a project plan and finish in a certain amount of time. You have to run experiments for a long time to work out if what you're doing is truly having an impact. And you've got to work out how to create feedback loops so that your machine learning improves and continues to improve over time.”

Rich Chetwynd

OneLogin Product Manager, Developer Experience

SEE FULL QUOTE

The black box problem limiting AI adoption

The black box problem limiting AI adoption

WATCH VIDEO

"By bringing in more intelligence, we can get closer to continuous authentication..."

“There's also a lot of invisible things that happen behind the scenes that we could do to improve the user experience. Those invisible things use the notion of AI, or behavioral analytics, to understand who the user is and what they're doing. By bringing in more intelligence, we can get closer to an era of continuous authentication. All of this is predicated on us bringing in more and more signals about the user, about the device they're on, and what they've done in the past. And all of this is going to help us reduce the inconvenience that these users face today.”

Jeff Broberg

Sr. Director, Product Management

SEE FULL QUOTE

"What AI/ML effectively means is that you're building a black box…"

“What AI/ML effectively means is that you're building a black box and putting a bunch of information in, and whatever the black box says is what goes. Customers aren't ready for that yet, because everybody needs to know why the answer happened. With machine learning, you don't necessarily know why; you just know the answer. It can make things more complicated. But over time, as the algorithms become more refined and more mature, people will have increased trust in that black box and questions of why a decision was made will start to dissipate.”

Rich Chetwynd

OneLogin Product Manager, Developer Experience

SEE FULL QUOTE