AI agents are pouring into companies, and most security tools were never built to watch them. Straiker wants to fix that. The startup has just raised $64mn to police a workforce that does not clock in.
The Series A was revealed first by Axios. It takes the Mountain View company’s total funding to $85mn. Marathon Management Partners led the round. Citi Ventures, Illuminate Financial and Workday Ventures joined, with earlier backers Bain Capital Ventures and Lightspeed returning.
Securing a workforce that never sleeps #
Straiker calls itself the agentic security company, and the pitch is simple. AI agents now book travel, write code and run back-office tasks. They act on their own across company systems. That independence is useful and risky in equal measure.
Old security tools assume software follows fixed rules. Agents do not. They reason on the fly and take actions no one scripted. That leaves a gap traditional controls cannot cover, the company said in its announcement.
The scale of the shift is the selling point. Research firm IDC expects more than a billion AI agents inside enterprises by 2029. That is around 40 times the number in 2025. No corporate workforce has ever grown that fast.
What can go wrong #
The threat is not theoretical. This spring, Reuters reported a striking case. Attackers talked Meta’s AI support agent into changing account emails and skipping two-factor checks. They hijacked more than 20,000 Instagram accounts without ever breaching Meta’s core systems.
Straiker’s own research arm, STAR Labs, found similar weak spots in testing. It said 36% of successful attacks on coding agents led to remote code execution. Another 91% of attacks on productivity agents ended in quiet data theft, with no malware and no stolen passwords.
The company’s platform tries to close those holes in three steps. It finds the agents already running in a business. It stress-tests them for flaws before launch. Then it watches them in real time to stop attacks as they unfold.
Pedigree and traction #
The founders come from the security world. Chief executive Ankur Shah previously ran Palo Alto Networks’ Prisma Cloud business. Chief technology officer Sreenath Kurupati led AI and security research at Akamai, after it bought his fraud-detection startup Cyberfend.
Since launching in 2025, Straiker has sold to frontier AI labs and Fortune 500 firms. It says run-rate revenue has grown more than 15-fold in under a year. “Demand is outpacing anything we forecast,” Shah said. The cash will go into the product, the STAR Labs research team and overseas expansion. Marathon’s Gokul Rajaram is joining the board.
A crowded new frontier #
Straiker is not alone in spotting the opening. A rush of startups now sells tools to test, govern and secure AI agents. They cover everything from non-human identity to agent access governance. Investors are betting that as agents get the keys to real systems, someone has to hold them to account.
The risk for the category is that big security vendors fold the same features into their suites. Straiker’s bet is that agents are different enough to need a dedicated guard. If IDC’s numbers are anywhere close, the market it is chasing will be huge. So will the cost of getting it wrong.
Get the TNW newsletter #
Get the most important tech news in your inbox each week.