AI and machine learning are the hot new features in tech right now, from drawings to data mining. You’ve probably seen AI offerings cropping up in your IAM products, including the upcoming Okta AI features for both workforce and CIAM use cases. But can they help you keep your users’ identities safer? That they can.
Bots compromise security, pollute customer intelligence, and degrade the user experience. They’re also hard to track, at least for humans. Machine learning, however, can draw data from millions of logins to catch bots in the act. AI has a high success rate with blocking bots, proactively detecting and blocking up to 80% of automated logins. AI can also quickly analyze every detail of a login to make sure that humans are still getting in seamlessly – analysis showed that less than 1% of bot-blocking challenges are shown to legitimate users.
For example, Okta uses ML to block “toll fraud”, or the use of bogus SMS requests to drive profits for fraudsters. Blocking toll fraud keeps the cost of SMS verification down for those companies that need to use it. This blocking both stops the attack and has a deterrent effect on attackers – convincing bad actors to move to weaker systems.
Starting in 2024, Okta AI will have additional security benefits beyond just keeping botted logins at bay. Policy Recommender will help organizations adapt their security policies in real time with recommendations based on your organization’s current configurations and best practices drawn from the Okta ecosystem.
AI isn’t just for administrators, though – users can also benefit from AI making user inputs more flexible and intuitive. Okta’s NLP (natural language processing) model lets users make access requests in their own words. With NLP, users can focus on their work rather than trying to match the request format for Okta requests. Auditors will also be able to leverage NLP to carry out log queries in natural language, helping them find and interpret log events.
With AI, you’ll be able to use your Okta data to increase security, make more effective governance decisions, and manage bots with less effort.