BNB Chain says its latest consensus-related upgrades have brought the network to sub-second finality, with the project pointing to BEP-590 and BEP-648 as key steps in the process. In an official blog post, the network said production finality has reached about 0.65 seconds after a sequence of upgrades.

The company described BEP-590 as an update that extended voting to recent ancestor blocks, a change it said was meant to absorb network jitter and improve finality speed. It said that approach brought average finality down to roughly 0.9 seconds.

BNB Chain 🤝 Sub-second finality

Apps, bridges, exchanges, and users now get settlement certainty in ~0.65 seconds, enabling faster deposits, cross-chain transfers, and near-instant confirmations.

Nearly 70x faster with zero EVM migration.

Here’s how BSC got there 🧵👇 pic.twitter.com/9D7plnCeQg

— BNB Chain (@BNBCHAIN) July 30, 2026

BNB Chain then pointed to BEP-648, which it said added an in-memory vote pool that aggregates votes in real time. According to the network, finality is confirmed at about 0.65 seconds once a two-thirds quorum is reached.

A sequence of upgrades, not a single switch

The blog framed the improvement as the result of a series of BEP upgrades rather than one isolated change. BNB Chain said the move from a 45-second finality model to around 0.65 seconds came through sequential updates, with BEP-590 and BEP-648 among the most recent steps highlighted publicly.

The sub-second result was presented as the outcome of layered consensus changes, not a standalone upgrade with an immediate effect on its own.

BNB Chain’s public thread also linked to the blog for further detail, but the core claim remains the same: the network says it has crossed into sub-second finality on production systems.

The post BNB Chain says BEP-590 helped push finality below one second appeared first on The Cryptocurrency Post.

NEAR Protocol is shifting its infrastructure toward the “agentic economy,” integrating confidential execution and intent-based systems to support autonomous AI agents. The latest updates, detailed in the NEAR official roadmap, highlight a focus on enabling machines to act as economic participants through private coordination and universal liquidity.

The core of this transition is Confidential Intents, a feature designed to provide privacy for cross-chain trading and payments. Unlike fully transparent decentralized finance (DeFi) protocols, which expose transaction details to bots and observers, Confidential Intents utilizes private shards to mask transaction size, direction, and asset pairs by default. The protocol aims to combine the privacy levels of specialized chains with the composability of public networks.

Adoption Metrics and Private Volume

Data from the second quarter indicates a significant shift toward private transaction execution within the ecosystem. According to an analysis by Nansen, Confidential Intents’ Total Value Locked (TVL) surpassed $30 million. Furthermore, the firm reported that 42% of the swap volume on the near.com interface was executed privately by default during this period.

4/ NEAR is also extending its cross-chain thesis into privacy.

Confidential Intents hides transaction size, direction, timing, and asset pairs while supporting selective disclosure.

By Q2:

→ Confidential Intents TVL exceeded $30M
→ 42% of https://t.co/POsLNajW9A swap volume…

— Nansen 🧭 (@nansen_ai) July 29, 2026

The integration of intents—a system where users specify a desired outcome rather than a sequence of technical steps—has expanded beyond native protocol tools. Third-party platforms including SimpleSwap and Ledger have recently incorporated NEAR-powered intent systems, facilitating swaps across more than 30 different blockchains.

Building Infrastructure for AI Agents

NEAR’s objective is to provide an “open stack” for AI agents that require more than just a large language model to function as economic actors. According to the protocol’s development team, an effective agentic economy requires several specialized capabilities, including:

• Confidential execution: Allowing agents to process sensitive data and financial moves without public exposure.
• Settlement speed: Infrastructure capable of handling machine-driven transactions at a micro scale.
• Universal liquidity: Enabling agents to access assets across multiple chains without fragmented user experiences.
• Private inference: Ensuring that the data processed by AI models remains secure.

The project characterizes Confidential Intents as a solution to the tension between high-speed execution and the liability of on-chain transparency. By running these flows on a private shard rather than relying solely on client-side zero-knowledge proofs, the protocol claims to maintain a standard user experience while preventing the leakage of trade positions.

Future Strategic Focus

The roadmap through 2026 continues to emphasize the convergence of AI and blockchain settlement. Beyond current swap and payment features, the protocol is working toward “Universal Send,” a feature intended to allow confidential payments to any user on any chain without exposing wallet history.

While the infrastructure for these agent-led transactions is scaling, with NEAR Intents cumulative volume reportedly reaching $19 billion in late May, the long-term success of the “agentic economy” depends on the developer adoption of these private execution layers and the continued integration of secure public-sector AI initiatives.

The post NEAR Integrates Confidential Intents to Anchor Emerging Agentic Economy appeared first on The Cryptocurrency Post.

NEAR Protocol has detailed a comprehensive framework for the role of digital assets in an economy dominated by autonomous agents, introducing what it describes as an “AI money” thesis. The protocol is positioning its native infrastructure as the coordination layer for “sovereign AI,” a stack designed to allow AI agents to operate with independent identity, private execution, and native settlement capabilities.

According to official communications from NEAR Protocol, the new economic model for AI agents shifts the function of money into four primary roles. It serves as a store of value to ensure network sustainability, a settlement asset for high-frequency micro-transactions between agents, a bonding standard for security, and a metering unit used to price inference and compute tasks.

Money has to work differently in an agent economy:

✦ Store of value → network sustainability: an asset tied to the security and economic coordination of the network it supports ⁰✦ Medium of exchange → settlement asset: value clearing with finality, cheap enough for…

— NEAR Protocol (@NEARProtocol) July 26, 2026

Infrastructure for Autonomous Agents

The push toward sovereign AI infrastructure focuses on moving beyond open-source models to create a decentralized environment where agents can function without central intermediaries. NEAR has identified several “sovereign agent” requirements that it intends to support through its protocol, including:

• Identity and Verifiability: Procedures to ensure agent authenticity and the ability to verify actions on-chain.
• Private Inference and Confidential Execution: Technologies that allow AI models to process data without exposing sensitive information.
• Open Settlement Rails: Financial infrastructure that allows agents to trade and move value autonomously.
• Liquidity and Coordination: Mechanisms to facilitate interactions between millions of independent digital entities.

The protocol describes itself as a holistic stack that integrates the base blockchain layer with “Intents” and “NEAR AI” to act as a coordinator for this open AI ecosystem. This positioning is detailed in a Q2 2026 report, which frames NEAR’s evolution not as a pivot, but as a return to its founders’ original focus on distributed training and compute systems.

Economic Sustainability and Token Burn

A central component of this infrastructure is the “Intents” layer, which is already impacting the protocol’s tokenomics. Current estimates suggest that the burn mechanism associated with Intents volume is a primary driver of network sustainability. According to some projections, net buyback pressure could exceed total token issuance if daily volume on the Intents layer reaches approximately $175 million, a figure the team views as a milestone for the agent-to-agent economy.

Despite the growth in this sector, the protocol acknowledges significant competition. While NEAR is scaling its infrastructure through milestones such as Multi-Party Computation (MPC) node expansion, its agent framework faces competition from other industry standards. Practical adoption for enterprise AI remains in the proof-of-concept stage, with long-term validation depending on repeatable commercial contracts and the velocity of “post-human” or agent-driven transaction volume.

The success of the “AI money” thesis will likely depend on the protocol’s ability to differentiate its settlement integration as autonomous workloads increase.

The post NEAR Protocol Outlines AI Money Thesis and Sovereign Agent Infrastructure appeared first on The Cryptocurrency Post.