Every message your AI sends, provable.

The scarce thing in 2026 is not the ability to send a message across many channels. It is the ability to prove afterwards who sent it — a human or an agent — on whose authority, under what policy, and whether it disclosed itself as AI.

What Kanvo is building

Category: Attestation infrastructure for agent-originated messaging.

Permission slips before. Receipts after.

Kanvo gives AI agents a permission slip before they speak, and a tamper-resistant receipt after they speak.

The problem

A company's AI assistant messages a customer: “Hi, your order arrives tomorrow.” Later, someone asks who sent that — and whether the company ever gave the assistant permission to say it. Today there is usually no answer. The log records that a message went out, and nothing about who was behind it.

What the receipt says

Companies already do, and they should. But an archive is a photograph of finished homework: it proves the homework exists. It does not prove who did it, or who said they could. Kanvo records that while the message is being sent, which is the only moment the information exists at all.

Why not just save the messages?

Companies already do, and they should. But an archive is a photograph of finished homework: it proves the homework exists. It does not prove who did it, or who said they could. Kanvo records that while the message is being sent, which is the only moment the information exists at all.

What that looks like

You may answer customers about shoe orders.You may not offer a refund over $500 without a human approving it.

Where Kanvo sits

AI agents reach messaging apps through Kanvo rather than around it. Standing in the middle is what makes the rest possible: before the message goes out, Kanvo asks the agent for its permission slip; after it goes out, Kanvo writes the signed receipt.

Why a business pays for this

Suppose an AI tells a customer “we guarantee you’ll earn 20% on this investment.” No company can answer that with “the robot did it.” Someone will ask who authorised the agent, what it was permitted to say, whether a human approved the message, and which rules it was following. Kanvo exists to answer those four questi…

A window opened on specific days

The argument is not that AI is important. It is that four dated shifts landed inside twelve months and together created an obligation businesses have acquired and cannot currently meet.

The obligation is already live

EU AI Act Article 50 became enforceable on 2 August 2026. Interactive AI must disclose itself and synthetic content must be machine-readably marked, with penalties reaching EUR 15M (roughly $16M) or 3% of worldwide turnover — and it binds any deployer whose AI output reaches the EU, not only EU companies.1,2,3

The primitives got standardised

Between October 2025 and February 2026 agent identity acquired a wire format: RFC 9421 message signatures over Cloudflare's Web Bot Auth draft, Visa's Trusted Agent Protocol, and AP2's signed mandates. Building this in 2026 is assembly rather than invention — and that was not true a year ago.8,9,10,11

The agents are already sending

Gartner expects 40% of enterprise applications to carry task-specific agents by the end of 2026, against agentic AI spending of roughly $202B. The messages are going out now; the record of who authorised them is not being kept.27,65

Every application is growing an agent of its own

The same Gartner figure read from the other side: if 40% of enterprise applications carry a task-specific agent by the end of 2026, then a working person is about to have a dozen assistants, each behind a different tab, each with its own history and its own idea of who they are. That is the fragmentation messaging spe…

The buyer already has the budget

Regulators collected more than $2B from over 100 firms for unarchived business messaging. The sweep ended but the recordkeeping rules did not change, and FINRA now surfaces the same violations in routine exams and has barred individuals personally.16,17,18

Why this company, and not a competitor

Every claim below is either verifiable in the codebase or carries a source. A company selling proof cannot ask to be taken on trust about itself.

Already on both sides of the wire

Attestation can only happen in the send path, by the system holding the delegation. Kanvo runs an MCP server that external agents call, delivers into ten channels, and signs in between — the envelope is minted after delivery, over the content the recipient actually received. Inbox platforms hold one side, agent-toolin…