Confidential Transfers

Ripple used a developer conference in Seoul, South Korea to detail a set of XRP Ledger upgrades built for institutions and autonomous software, including confidential transfers, transaction batching and dedicated infrastructure for AI agents. The moves push the network beyond simple payments toward workflows that banks and automated systems can use without exposing sensitive transaction data.

• Ripple plans confidential transfers using zero-knowledge proofs so institutions can hide transaction details while still proving validity to regulators.
• AI agents have already generated more than 11 million transactions on the XRP Ledger, according to Ripple engineering lead Ayo Akinyele.
• The ledger settles transactions in roughly three to five seconds, a speed Ripple says suits machine-to-machine payments.

• 11M+ AI agent-driven transactions already recorded on the XRP Ledger
• 3-5s settlement time Ripple cites for machine-to-machine payments
• 3 major upgrade categories unveiled: privacy, batching and AI infrastructure

The XRP Ledger, the blockchain network used mainly for cross-border payments, has operated with full transaction transparency since launch. That openness has long been a sticking point for banks and asset managers who cannot publish the details of complex trades or settlement flows on a public ledger, even when they want the speed and finality blockchain settlement offers.

Ripple Builds Confidential Transfers on Zero-Knowledge Proofs

Ripple’s answer is a confidential-transfer feature built on zero-knowledge proofs, a cryptographic method that lets one party prove a statement is true without revealing the underlying data. Applied to XRPL, it would let an institution shield the specifics of a transaction while still allowing the network and regulators to verify it followed the rules, according to a presentation posted on X by an account covering the conference.

Alongside privacy, Ripple is adding transaction batching, which bundles multiple operations into a single all-or-nothing execution. The company is pairing that with formal verification and adversarial testing, engineering practices meant to catch flaws before code reaches production, as it positions XRPL for the kind of multi-step settlement logic institutional trading desks require.

Akinyele Says AI Agents Have Already Moved 11 Million Transactions

Ripple’s Senior Director of Engineering, Ayo Akinyele, told the Seoul audience that AI agents have already driven substantial activity on XRPL, with related transaction volume topping 11 million. That figure predates any of the infrastructure Ripple announced at the conference, suggesting agent-driven usage is forming organically ahead of dedicated tooling.

AI agents have already generated substantial activity on XRPL, with related network volume surpassing 11 million transactions.

Ayo Akinyele, Senior Director of Engineering, Ripple

To support that growth, Ripple is building verifiable identities, spending controls and secure smart wallets that can hold both XRP and stablecoins. The aim is letting software agents transact on their own within limits set by their human operators.

That design mirrors a broader push across crypto infrastructure to let autonomous systems pay for services directly, an area covered in our earlier reporting on AI agent spending tools. Ripple’s three-to-five-second settlement window compares favorably with the days traditional cross-border rails can take, which is the pitch Ripple is making for machine-to-machine commerce specifically.

BIS Paper Already Floated XRPL for Official Statistics

The Seoul announcements, first detailed by CryptoPotato, follow a working paper from the Bank for International Settlements that explored using XRPL to verify official statistics on-chain without disrupting existing data systems. That prior research gives Ripple’s institutional pitch some external validation beyond its own marketing.

None of the upgrades described in Seoul have shipped yet, and Ripple did not give a timeline for mainnet deployment of confidential transfers, batching or the AI agent wallet infrastructure. Questions around how regulators will treat confidential, zero-knowledge-shielded transactions on a public ledger remain open, an issue that echoes debates playing out elsewhere in crypto infrastructure, including the SEC’s proposed crypto custody rule for investment advisers.

The CCS read. We read this as Ripple hedging two bets at once: privacy tooling to court banks wary of public ledgers, and agent infrastructure to capture a transaction category that did not exist when XRPL launched. The 11 million agent transactions already on the network matter more to Ripple’s roadmap than any price action, since they show demand arriving before the infrastructure does.

Ripple has not set a date for mainnet rollout of confidential transfers or the AI agent wallet stack, leaving institutional adopters to wait for a formal deployment timeline before they can test the privacy and batching features against real settlement workflows.