AI Liability Infrastructure

The trust layer agentic AI is missing.

Autonomous agents are executing transactions, moving money, and making decisions at tera-scale. Teracor keeps the record: a passive, tamper-evident audit trail of every action they take.

Get in touch The problem
AB 316
California
in effect
Jan 2026
···
One defense
is gone

If your product is built on Claude, GPT, or any foundation model and your agent acts outside its defined scope, AB 316 removes the argument that the AI acted on its own. It applies to anyone who developed, modified, or used it: your model provider, your vendor, and you. In practice, you are the first party asked to prove what happened. Teracor is the receipt that proves what your agent was configured to do, what it actually did, and when.

Nothing here is legal advice. The litigation logic follows from the statute.

01: The Problem

"The AI did it" is no longer a defense.

California AB 316 went into effect January 2026. It forecloses the argument that enterprises can fully deflect responsibility onto the AI system itself. Each party in the deployment chain loses the ability to point at the AI and step back. Enterprises deploying agentic AI in regulated contexts need to understand where they stand.

The EU AI Act is a separate regime with a different mechanism. AB 316 governs liability after harm occurs in California. The AI Act imposes compliance obligations on AI systems placed on the European market, whether or not anything has gone wrong. Article 50 transparency obligations took effect August 2, 2026, alongside penalty powers and market surveillance. High-risk obligations under Annex III were deferred by the Digital Omnibus to December 2, 2027, and to August 2, 2028 for AI embedded in regulated products. A company can be fully compliant with one regime and exposed under the other.

Carriers are repositioning against both. New ISO exclusion forms effective January 2026 give carriers the option to exclude AI-related claims, with adoption accelerating at renewal. Gallagher Re reports that courts and regulators are treating AI as a tool rather than an independent legal actor, which places responsibility for its outputs on the deploying organization, while standard vendor contracts cap liability at twelve months of fees and offer no performance warranties. The deployer carries the exposure without visibility into the model.

When a claim, a regulator, or a courtroom asks what an agent actually did, most enterprises have engineering logs generated by the same systems whose conduct is in question. That evidence is self-attested. It carries no independent weight.

The market is responding, but the critical gap remains unaddressed.

California
AB 316
Jan '26
In effect. Autonomous harm
defense stripped across
the deployment chain
European Union
AI Act, Article 50
Aug '26
In effect. Transparency
obligations and penalty
powers now active
European Union
AI Act, Annex III
Dec '27
High-risk obligations apply.
Deferred from Aug 2026
by the Digital Omnibus

02: What We're Building

A passive, tamper-evident audit trail for AI agents in production.

Teracor observes everything and blocks nothing. It records every agent action as a cryptographically verifiable receipt, without touching your agent code. Its independence from the decision chain is structural, not a feature.

Certification frameworks validate agents before deployment. Governance and enforcement tools decide what agents can do. Teracor records what agents did. A system that makes decisions about agent actions becomes a party to those actions. It cannot serve as an independent witness. Teracor can, because it holds no authority over the outcome it records.

The receipt that matters in court is the one generated by a system that had no stake in the outcome.

The platform is live. We are in design partner conversations with CTOs and VP Engineering at SaaS companies built on foundation models, and with legal and compliance leaders in fintech and insurtech navigating AB 316 and the EU AI Act. If your agent acts and you can't prove what it did, we want to talk.

Live Now
Audit Trail SaaS
A tamper-evident record of every agent action, with rule evaluation and a compliance dashboard, for enterprises deploying AI agents under AB 316 and the EU AI Act. The record is generated by a system outside the decision chain. That is the only version of the record that holds up. Design partners get full platform access.
Roadmap
Neutral Data Layer
The sealed record becomes infrastructure other parties can build on: carriers, counterparties, and courts that need a version of events generated before money was at stake, by a party with no position in the outcome. Teracor sells no coverage and pays no claims. Details available in conversation.

03: What Teracor Is

Teracor is

  • A passive observation layer. It watches agent traffic and writes the record. It takes no action of its own.
  • A tamper-evident record. Every receipt is cryptographically verifiable. Alteration is detectable by design.
  • Independent of the decision chain. Teracor holds no authority over any agent action, which is what makes its record credible as evidence.
  • Deployable without code changes. Teracor captures the decision history of your agents without touching your agent code.

Teracor is not

  • An enforcement layer. Teracor never sits in the execution path. There is no kill switch and no blocked state, by design.
  • A governance tool. Teracor takes no position on what your agents should be allowed to do. It records what they did.
  • Developer observability. Latency traces and token counts serve engineers. Teracor serves the moment a regulator, a carrier, or a courtroom asks for proof.
  • An insurance product. Teracor sells no coverage, underwrites no risk, and pays no claims. It is compensated identically whether a claim pays or does not. Independence is permanent, at every phase.

04: Connect

Your AI agents are making decisions.
Can you prove what they did?
Let's talk.

Teracor is looking for design partners: SaaS platforms built on foundation models first, plus fintech and insurtech teams deploying agents under AB 316. Partnerships start with an LOI, a sandbox integration, and full platform access at no cost. The earlier the conversation, the more we build around your specific exposure.

partners@teracor.ai