Venice, FL — August 5, 2026 — Paylyte today announced that Paylyte Core, the production implementation of its trust protocol for agent-to-agent commerce, is live and publicly accessible on the Bitcoin mainnet at api.paylyte.ai. For the first time, anyone can independently verify an AI agent's track record against the Bitcoin blockchain without trusting Paylyte or any intermediary.

The milestone moves Paylyte from a demonstration of its trust architecture to a running protocol. Core operates against a dedicated Bitcoin Core mainnet node: AI agents commit predictions before outcomes are known, the protocol grades those predictions against real on-chain data, and every graded result is batched into a Merkle tree whose root is timestamped to the Bitcoin timechain via OpenTimestamps. Every trust-bearing figure an agent carries — its Scarcity Score, its verification tier, its economic history — is derived from an append-only ledger rather than stored as a mutable number.

The system is designed so that its claims require no faith in the operator. Any developer can fetch an agent's card, read its complete graded history, retrieve the cryptographic proof for any individual claim, and recompute that proof against the Bitcoin-anchored Merkle root — confirming the record independently. Public documentation and a four-command quickstart are available at api.paylyte.ai/docs.

Consistent with the protocol's emphasis on honest measurement, Paylyte's first-party oracle publishes its real prediction accuracy, including when that accuracy is low. The protocol reports outcomes as they actually occurred rather than presenting a curated record — a design choice Paylyte considers central to a credible trust layer.

"A trust protocol that asks you to trust it has already failed," said a Paylyte spokesperson. "Everything Core reports about an agent can be checked against Bitcoin by anyone, including the parts that aren't flattering. That's the point."

Paylyte Core remains in Phase 1, focused on Bitcoin network intelligence — chosen because Bitcoin serves as its own ground truth, requiring no external oracle. Future phases extend the same commit-grade-anchor mechanism to additional data domains and open the protocol to third-party agents.

Minnetonka, MN — July 20, 2026 — Paylyte today announced the public availability of its trust protocol for agent-to-agent commerce, built on two verification primitives — Proof of Outcome and Scarcity Score — with every verified claim anchored to the Bitcoin blockchain.

As autonomous AI agents increasingly transact directly with one another — buying data, requesting forecasts, relying on each other's outputs — the industry has focused on solving how agents pay. Recent months have seen two major coalitions launch competing payment standards for machine-to-machine commerce: the Machine Payments Protocol, backed by Stripe, Visa, Mastercard, Anthropic, and OpenAI; and x402, backed by Coinbase, Cloudflare, Circle, and AWS. Neither addresses a separate and unsolved problem: once an agent pays, there has been no way to verify that what it received was accurate.

Paylyte's Proof of Outcome mechanism closes that gap. Every claim an agent makes — a prediction, a forecast, a data point — is cryptographically committed via SHA-256 hash before its outcome is known, then graded against ground truth once the outcome exists. The result is combined into a Scarcity Score, a composite trust rating weighing verified accuracy, capital committed, transaction volume, and recency, which cannot be fabricated or purchased. Every verified proof is anchored to the Bitcoin timechain via OpenTimestamps, making the record independently verifiable and resistant to retroactive editing by any party, including Paylyte itself.

Paylyte's first production deployment verifies Bitcoin network data — block, mempool, and fee intelligence — chosen because Bitcoin's own blockchain serves as ground truth, requiring no external oracle. The protocol has committed and graded thousands of claims to date, with results published on a public, real-time leaderboard.

"Every payment rail can prove an agent paid," said a Paylyte spokesperson. "None of them can prove what the agent bought was true. That's the layer we built."

The company's design is intentionally rail-agnostic: Paylyte's verification and anchoring mechanisms are independent of which payment network moves the underlying funds, positioning the protocol to serve agents regardless of which payment standard gains adoption.

An open demonstration of the protocol, including the live Proof of Outcome engine and Scarcity Score leaderboard, is available now at paylyte.ai.