Tide's MCP Server Tackles AI Security by Rethinking Authority Distribution

Tide's latest MCP server enables coding agents to weave security into applications through what the company calls "emergent authority" — a paradigm where power to access systems and data is generated on demand rather than stored permanently in one location. Co-founder Michael Loewy explains that defenders have always faced an asymmetric challenge: they must be flawless, while attackers need only succeed once.

"Before AI even came into the developer consciousness, a developer or a platform owner needed to be perfect all the time — perfectly patched, free of errors, free of bugs — and an attacker only needed to be right once to get in," Loewy tells The New Stack. "It's virtually impossible, which is why we're seeing breaches in the news every day, and the biggest companies in the world getting breached. It could be one poorly written line of code, one misconfiguration, and then you're done."

The arrival of AI has intensified these dangers. Loewy cautions that AI systems are producing vast quantities of unvetted, error-laden code while simultaneously becoming adept at finding and weaponizing security flaws. AI agents themselves are now writing software and executing privileged operations, expanding the threat surface.

Attackers typically breach systems by circumventing authentication and authorization controls. However, Loewy and co-founder Ben Waters focus on a different concern: the authority that agents wield once they gain entry. Authority encompasses the ability to determine which users or systems access which resources, who decrypts which information, and who grants permissions. This power traditionally resides in admin credentials, root keys, identity providers, and service accounts — and someone always holds the keys.

The flaws of the all-eggs, one-basket approach

Most organizations concentrate application secrets, user credentials, and access permissions in a single repository, then layer defenses around it: firewalls, key vaults, multi-factor authentication, and endpoint detection tools. Ben Waters argues this foundation is inherently broken.

Even if your code is perfectly free of bugs, it's sitting in someone else's cloud on someone else's operating system. There's this whole suite of dependencies where the code and everything needs to be perfectly patched all the time, everywhere, and configured correctly for your security to be perfect. That root paradigm is flawed. Tide seeks to overcome that by using cryptography to compute authority in pieces so that it can't be reassembled.

Ben Waters

Tide's model, called "emergent authority," distributes power so that it only materializes when identity, policy, context, and intent all align — then vanishes. The goal is that even if an attacker gains root access to your server, there is nothing to steal because authority has been architecturally dispersed elsewhere.

From IoT middleware to distributed authority by default

Tide began roughly a decade ago as an analytics platform for Internet of Things devices, connecting brands with consumers' health gadgets, smart home systems, and wearables. Regulated customers demanded assurance that breaches could not expose sensitive data, trigger GDPR penalties, or damage reputations. The security infrastructure Tide built for itself became the product.

That infrastructure is now TideCloak, a developer product positioned at the identity and access management layer. It handles authentication, authorization, end-to-end encryption, and governance atop Tide's Cybersecurity Fabric, ensuring consequential authority stays out of reach. The implementation is fully inspectable through open-source repositories on GitHub, and Tide published a white paper on emergent authority using a chatbot to clarify the underlying concepts and cryptography.

Let your coding agent do the integration

Rather than requiring developers to wade through extensive documentation, Tide designed its security system for adoption through AI agents. Loewy notes that "We needed this security apparatus or infrastructure that we've created to be something that can be seamlessly integrated into an existing project or best practice in a greenfield project, in a way where the developer doesn't need to read loads of documentation."

On Monday, Tide unveiled Raziel, an MCP server named for the archangel of secrets. Raziel equips AI assistants with expertise in Tide authentication, threshold cryptography, end-to-end encryption, and governance. Developers can point their agent at Raziel, which then recommends self-hosted or managed deployment options and generates verified playbooks for integration. A TideCloak quickstart guide addresses greenfield scenarios for teams starting fresh.

Unlike conventional security tools that hunt for entry points, one of Raziel's capabilities maps your blast radius by assuming the attacker is already inside. Loewy explains the shift in perspective: "Where a typical cybersecurity scanner tries to find vulnerabilities — how someone's going to get into your platform — we've started the opposite way. We've said: they're already in. Here's what they're going to find. Here's where they can impersonate any user. Here's where they can assign themselves access to whatever they want."

Tide has also secured a partnership with a Lloyd's of London underwriter, allowing organizations that deploy TideCloak to access preferential cyber insurance rates. Loewy notes this represents "a rare case of a security claim being backed by capital."

Coordination at scale, without custody at scale

The team envisions Tide eventually becoming infrastructure similar to DNS — owned by no one yet benefiting everyone. The implications extend beyond breach prevention. Whenever an AI agent, contractor, partner, or new vendor needs to perform critical operations, you must grant them significant power. This trust requirement constrains how much responsibility you can safely delegate or automate.

The point isn't just safer systems. Once nobody has to hold dangerous power to get something done, you can delegate real responsibility to an agent, a contractor, or a five-person startup without creating a dangerous insider. That's what lets developers ship with AI at full speed. It's coordination at scale without custody at scale.

Michael Loewy

If this approach proves viable at scale, Tide's most significant impact may transcend cybersecurity itself, fundamentally changing the economics of delegation and expanding what small teams working alongside AI agents can safely accomplish.