By: Jake Smiths

In today’s digital economy, identity has become the new perimeter. Organizations must manage who has access to what across an expanding landscape of applications, cloud services, and AI tools. Despite years of investment in security programs, enterprises still face blind spots, fragmented identity stacks, and operational complexity.

To explore the challenges and the role AI is playing, TV reporter Jane King had an interview with Yossi Barishev, founder and CEO of a stealth identity security company, about the current state of identity security. 

Identity: The Hidden Bottleneck

Barishev’s career began in cyber incident response, where he saw how identity gaps hindered remediation. “When you do cyber incident response, obviously, you see all the faults and all the flaws in an organization kind of flowing upwards and the biggest bottleneck consistently was always identity. Figuring out what are the compromised identities? How do we get to them? How do we remediate the attacker’s footprint in the environment? Scouring after that data, and data is a big gap, context,” said Barishev.

Later, leading security teams, he noticed the same challenge repeated: “I saw my teams constantly scaring after identity-related data, identity-related context, who has access to what, why, who approved that… all of these moments kind of made me realize that there’s definitely a major bottleneck right now in identity and at the same time now more than ever, identity is the new perimeter everybody sets,” he added.

Defining identity security in its simplest form, Barishev explained, “Identity is the practice of making sure that the right people have the right keys to the right doors, right? The keys being their credentials, so to speak, and the doors being whatever systems exist in an organization,” said Barishev.

The Complexity of Modern Identity

The explosion of SaaS and AI applications has increased identity management complexity. “Too many pains to deal with, too many data sources to deal with,” said Barishev, describing how companies have gone from a few identity solutions to fragmented stacks of seven to fifteen systems.

AI brings both promise and challenges. “A lot of things were overpromised about AI. If anything, it’s still perceived as this silver bullet magic wand… In reality, AI has its limitations. Its main limitation, by the way, is whatever you feed it with. If you feed it with junk, it will output junk,” said Barishev.

He offered a practical example: “If I’ll feed it the wrong context… it will give me an answer based on that and so if I give him a fragment of the truth, he will give me an incorrect answer and then I’ll have to act upon an incorrect conclusion,” said Barishev.

Blind Spots and Compliance Theater

Even with AI, enterprises often operate under the “illusion of coverage.” “There’s a lot of disconnected applications in any enterprise environment… these are the blind spots that people miss,” said Barishev.

Compliance audits sometimes become performative rather than meaningful. “It’s really hard to get this data in a real-time fashion. By the time you act upon it, the reality behind the changes has moved,” said Barishev.

Mergers and Acquisitions: A Compounding Challenge

Identity management becomes even more complex during mergers and acquisitions. “An M&A is just not a merger of two businesses. It’s a merger of two IT stacks, two pieces of infrastructure… Usually, what happens is those two companies end up being kind of separate, kind of together in this mishmash state. It gets ugly,” said Barishev.

Integration of identity systems often takes months, leaving gaps that adversaries could exploit.

Building a Foundation in Data

Barishev sees high-quality identity data as the foundation of any effective security program. “If I could spend just $1 on this problem, I would spend it on improving my identity-related data… Every single system you build on top of this data will either suffer from a garbage-in, garbage-out problem or it will have reliable and clean data to work,” said Barishev.

He predicts the next wave of solutions will provide more than visibility. “People really care about how we can have a centralized source of truth for all things identity to act upon in a meaningful way, a new foundation, so to speak, on top of which you can build your future,” said Barishev.

As enterprises expand their software footprint and adopt AI, identity security remains a critical bottleneck. Barishev’s perspective makes clear that without clean data and a unified approach, even the most advanced AI cannot solve the problem.