AI is a double-edged sword for cybersecurity. It is creating new attack categories. It is also making the old ones faster, cheaper, and harder to catch. And this was confirmed by two recent reports.
ReliaQuest's 2026 AI-powered cybercrime report and
FusionAuth's 2026 State of AI and Identity Report come at the problem from opposite directions but land on the same point. ReliaQuest found threat actors using AI to scale phishing, social engineering, malicious code, and credential harvesting. None of that is new. It's just cheaper and faster to run now.
FusionAuth's numbers show where that pressure actually lands inside organizations: 65% of respondents confirmed an AI identity-related security incident in the past year, and 88% said AI deployment is moving faster than their identity and security infrastructure can handle.
The pattern is obvious. Attackers are using AI to do more of what already worked, while organizations keep adding AI into their environments faster than they can govern who or what actually has access. And this is what is really the story for MSSPs - not a detection problem, but an access control problem, and that distinction matters because it points to a different kind of gap. When organizations expand AI use without governing agent access, verifying tenant isolation, or containing what happens when a token gets compromised, they're building the exposure themselves. No attacker required.
And that’s where the real MSSP work starts: checking identity posture, knowing machine identities, finding shadow AI use, validating tenant isolation, tuning detections, and making sure incident response plans actually account for AI-specific failures. It is also a conversation MSSPs should be having directly with clients: confidence is not the same as control. The organizations that feel most secure about AI may simply be the ones deploying it the fastest. That should raise some questions. Becasue then it comes down to access and visibility into what AI tools and agents are doing.
Market Pulse: Cybersecurity Deals, Funding, and Platform Shifts
Keeper Security addresses credential drift with Universal Secrets Sync: Keeper Security has added Universal Secrets Sync to KeeperPAM to help security teams keep cloud secrets from drifting out of sync. The new capability automatically pushes rotated secrets from Keeper Secrets Manager shared folders into AWS Secrets Manager, Azure Key Vault, and Google Cloud Secret Manager. That matters because secrets often get updated in one place but remain stale somewhere else in production, leaving teams with access failures, slow incident response, and credentials they may not be able to see or revoke. For MSSPs and security teams managing cloud and hybrid environments, this gives a cleaner way to keep secrets aligned across platforms without manual exports, custom scripts, or reconfiguration after every rotation.
Akamai secures the agentic web: Akamai has introduced an agentic security framework for its Bot & Agent Control portfolio, aimed at helping organizations verify and manage AI-driven interactions across browsers, bots, and agents. The framework connects identity, observability, trust, and edge enforcement into a real-time decisioning layer, with a Know Your Agent protocol designed to link AI agents to authorized human users and clarify identity, origin, and intent. Akamai says the framework is built around six areas, including verified agent identity, user-centric authentication, adaptive trust analysis, edge-based enforcement, content monetization, and traffic visibility.
Zscaler, Gigamon bring visibility to Zero Trust access: Gigamon has partnered with Zscaler to bring deeper application visibility into Zero Trust private access environments. The new integration combines Zscaler Private Access with Gigamon Application Metadata Intelligence, giving security teams more context into what happens after a user is granted access to a private application. That matters because ZTNA can control who connects, but teams still need visibility into user-to-application behavior, East-West traffic and possible lateral movement across hybrid cloud environments. For MSSPs, the integration points to a growing customer need around Zero Trust operations: validating least-privilege policies, spotting suspicious activity after access is approved, and feeding richer network-derived telemetry into SIEM, NDR and SOC workflows.
Nebulock raises $25 million Series A:
Nebulock has raised $25 million in Series A funding led by FirstMark, with participation from Bain Capital Ventures, Decibel, Zetta Venture Partners and Step Function. The company says the funding will support expansion of its AI-native contextual security platform, which is designed to help security teams hunt for threats across endpoint, identity, cloud, network and SaaS environments. Nebulock says its platform has already run more than 300 million agentic investigations and produced more than 4,000 high-confidence findings for customers, including cases involving malicious remote access, insider data movement, exposed credentials and risky browser extensions.
Have news to share or just want to connect? Reach anytime at [email protected].
