Thursday 20 August 2026

AI/Innovation·Pulse

> the signal in the ai noise
[ AI Infrastructure ]

The Billion-Dollar Bet That AI Agents Are Now a Security Problem of Their Own

AI & Innovation Pulse Editorial/4 min read/8 August 2026
A dark enterprise data centre corridor with glass walls, server racks, and glowing network diagrams suggesting autonomous digital agents moving through access paths.
Non-human identities are now the fastest-growing identity category inside large enterprises. The data centre is where the governance problem lives.

“Fortune 500 companies have seen the number of non-human digital identities inside their systems grow by nearly 500% in six months. The market has decided that securing AI agents is now one of the most urgent problems in enterprise technology.”

Fortune 500 companies have seen the number of non-human digital identities inside their systems grow by nearly 500% in six months. A billion-dollar acquisition in Israel's cybersecurity industry suggests the market has decided that securing AI agents — not just the humans who use them — is now one of the most urgent problems in enterprise technology.

A Deal That Signals a New Security Category

On 28 July 2026, data security company Cyera announced it had signed a letter of intent to acquire Oasis Security for approximately $1 billion, in a deal expected to be paid mostly in cash with the remainder in Cyera shares. The acquisition is Cyera's third in 2026 alone, following recent purchases of Index Ventures-backed Ryft and the sub-one-year-old Genie Security.

The strategic logic behind the deal is specific and increasingly urgent. Cyera's existing platform focuses on data security — discovering and classifying enterprise data, assessing its sensitivity, and governing who can access it. Oasis Security, founded in 2022 by Danny Brickman, a former head of cyber R&D for the Israel Defense Forces, and Amit Zimerman, specialises in an adjacent but distinct problem: managing and securing "non-human identities" — the machine accounts, service credentials, and increasingly, autonomous AI agents that now operate alongside human employees inside enterprise systems.

Why Non-Human Identity Has Become a Board-Level Issue

The scale of the problem is genuinely striking. The number of non-human identities inside Fortune 500 companies grew by nearly 500% in the six months preceding the announcement — making them the fastest-growing identity category inside large enterprises today. Gartner's forecasting is even more dramatic looking further out: the research firm predicts that by 2028, the average global Fortune 500 enterprise will be running more than 150,000 AI agents, up from fewer than 15 in 2025.

The core problem, as described by Oasis co-founder and CEO Danny Brickman, is structural. Traditional identity systems were built around human employees and slow-changing, standing permissions. AI agents behave completely differently: they can be spun up in seconds, act autonomously on a task, and be gone again within the hour. Cyera Chief Strategy Officer Jason Clark framed the shift succinctly: "every single agent is an identity" — meaning it needs to be tracked, governed, and controlled the same way a human user would, just at a radically different speed and scale.

What the Combined Platform Is Meant to Do

The stated ambition is to unify two capabilities traditionally managed separately: knowing what data exists and how sensitive it is, and knowing which identities — human, machine, or AI agent — can reach that data. Cyera CEO Yotam Segev put it directly: knowing your data isn't sufficient if you can't govern who or what touches it, and knowing your identities isn't sufficient without understanding the data context around what they can access.

Brickman, who will lead the combined engineering effort, has framed the ambition beyond simple restriction: the goal is to let enterprises "automate their most critical work without letting a single wrong decision interrupt the business" — positioning the technology as the governance layer that makes faster agent deployment safe at scale.

A Consolidating, Competitive Market

The deal doesn't exist in isolation. The non-human identity security space has seen a wave of consolidation through 2026: rival platform SailPoint has acquired Entro, and 1Password acquired Apono for a reported $250-300 million. KuppingerCole's Leadership Compass for the sector names nine "Overall Leaders," and every one is an established identity, privileged access management, or secrets management vendor — including CyberArk, Delinea, HashiCorp, and Microsoft — highlighting that deep capability in this space currently correlates with already owning foundational infrastructure. Analysts note neither Cyera nor Oasis is fundamentally a secrets management vendor, meaning enterprise buyers should understand exactly what dependencies the combined platform carries.

Cyera enters the deal from a position of considerable financial strength but not yet profitability: the five-year-old company recently raised $600 million at a $12 billion valuation and has surpassed $150 million in annual recurring revenue, having raised roughly $2.3 billion in total funding. Independent analysis has characterised the acquisition price as a deliberate premium bet on the "agentic AI" narrative continuing to accelerate.

What This Could Mean for Ireland

Ireland's position as European headquarters for a substantial share of the world's major technology and financial services companies means the non-human identity problem this acquisition addresses is not abstract for Irish enterprise. Multinational firms operating out of Dublin, Cork, and elsewhere are among the organisations most likely to already be deploying AI agents at scale within regulated data environments, making governance over what those agents can access a live compliance concern under both existing data protection law and the EU AI Act obligations Ireland's new AI Office is tasked with coordinating from August 2026. Irish cybersecurity and data governance firms may also find genuine commercial opportunity here, as large multinational clients look to implement non-human identity governance locally.

The Bottom Line

A billion-dollar acquisition between two Israeli cybersecurity companies might, on the surface, look like a niche industry transaction — but the underlying numbers it responds to tell a much bigger story: AI agents are proliferating inside major enterprises at a pace traditional identity security tools were never built to handle, and the market has now placed a very large financial bet that governing what those agents can see and do will be one of the defining enterprise security challenges of the next several years. Whether Cyera and Oasis successfully integrate two different platforms remains to be proven — but the strategic direction the deal signals looks set to become standard practice well beyond this one transaction.

AI & Innovation Pulse is an independent Irish digital title covering artificial intelligence and innovation. Every piece is written in-house, editorially independent, and verified.