verifiable

AI security startup Attestable has officially announced its establishment and completed a $20 million seed round financing. This round of financing was led by Jamin Ball of Altimeter Capital and Yonatan Mandelbaum of TLV Partners, with participation from institutions such as Halcyon Futures, Cerca Partners, and several investors.Attestable founder Yogi stated that as AI gradually enters critical infrastructure, national security, and large enterprise systems, how to verify the credibility of AI operations has become an important issue faced globally. The company aims to build a universal verification layer for cutting-edge AI laboratories, critical infrastructure, and national-level applications.According to reports, Attestable uses Zero-Knowledge Proof technology to shift trust in AI systems from data centers to mathematical verification mechanisms. This technology can prove that a certain approved model, model weights, input data, and operational strategies indeed generated specific outputs, while not disclosing model parameters or user privacy data, and without needing to rerun the model for verification.Attestable stated that its technology has currently achieved verification inference of the Meta Muse Glimmer 30B model on a single NVIDIA H100 GPU, reaching a speed of 85 tokens per second. The generated proof documents are small and possess quantum-resistant characteristics, allowing for rapid verification.In response, Ethereum co-founder Vitalik Buterin commented, stating that this achievement means the performance loss of Zero-Knowledge Proofs for large language models is approaching single-digit levels. He noted that the next challenge is to further reduce the performance overhead of technologies such as Fully Homomorphic Encryption (FHE).

According to research by CointelegraphMT, traditional financial markets are absorbing institutional capital at a pace that cryptocurrencies find hard to match by 2026. The main reason is that AI has clear and measurable returns, while cryptocurrencies currently lack a similar narrative.Data shows that the S&P 500 index rose only 3.5% after excluding AI stocks in 2026, while AI-related indices saw an increase of nearly 50%. The five major tech companies in the U.S. are expected to reach $72.5 billion in capital expenditures for AI infrastructure this year, with Nvidia's quarterly revenue reaching $81.6 billion.The research points out that AI spending can be directly validated through revenue, capital expenditures, and profit margins, while the value proposition of cryptocurrencies is difficult to quantify for traditional allocators. Currently, while the supply of stablecoins is at a historical high, more funds are flowing into tokenized government bonds rather than risk assets.Additionally, in May, the net outflow from U.S. spot Bitcoin ETFs was $2.3 billion, marking the worst single month of the year. However, long-term holders continue to buy in the over-the-counter market, with market makers like Wintermute reporting stable buying around $72,000. The research concludes that unless cryptocurrencies can provide a measurable and repeatable institutional-level return story similar to AI, they will be at a significant disadvantage in competing for the same institutional funds.

According to market news, Boundary Labs, a stablecoin startup led by Galaxy Ventures under Galaxy Digital, announced that it is preparing to launch an institutional stablecoin USBD and has completed a $2 million seed pre-financing.It was introduced that this round of financing also attracted participation from institutions such as First Block Capital and BlackWood. Boundary Labs was founded by former Deutsche Bank and Digital Currency Group executive Matthew Mezger. USBD will be deployed on the Ethereum network, focusing on the concept of "verifiable stablecoins," attempting to reduce the reliance of traditional stablecoins on off-chain audits and trust mechanisms through continuous on-chain verification of reserves, net asset value (NAV), and protocol operational status.The project team stated that USBD is designed with over-collateralization and hedging strategies to reduce market volatility risks and shift the stablecoin infrastructure from a "trust-driven" model to a "verifiable financial system." Additionally, Boundary plans to launch a staking token sUSBD for distributing protocol earnings under a delta-neutral DeFi strategy, but USBD itself does not provide yield functionality.The team indicated that the protocol is primarily aimed at asset management institutions, hedge funds, and family offices, and plans to officially launch the mainnet in early summer 2026.

Mastercard and Google jointly released the open standard framework Verifiable Intent, designed specifically for the "agent commerce" scenario where AI agents autonomously complete shopping transactions. This framework utilizes cryptographic technology to create tamper-proof authorization records when AI agents execute transactions on behalf of users, integrating identity, intent, and actions into a single privacy-protecting audit trail, addressing the issue where merchants and issuing institutions cannot confirm whether consumers truly authorized a transaction when AI agents place orders autonomously.The framework employs selective disclosure technology to ensure that parties only receive necessary information and integrates widely adopted industry standards such as those from the FIDO Alliance and EMVCo. Mastercard stated that Verifiable Intent will be integrated into its Agent Pay API in the coming months.Meanwhile, the cryptocurrency camp is also actively positioning itself. Coinbase CEO Brian Armstrong pointed out that while AI agents cannot open bank accounts, they can hold cryptocurrency wallets, indicating that blockchain infrastructure may be more suited to the future of the AI agent economy.

The founder of Venice, Erik Voorhees, posted on the X platform stating that the current privacy model of the Venice platform does not store prompts and response content on servers, and they will not be retained, viewed, or extracted, but this privacy cannot currently be proven.He stated that verifiable privacy protection has been on the roadmap since the project's inception, and this feature will be delivered soon.

In the report "2026 Digital Asset Outlook" published by Animoca Brands Research, 0G Labs CEO Michael Heinrich emphasized the importance of "verifiable AI." He pointed out that as the development of autonomous agents surpasses human regulatory capabilities, blind trust will lead to systemic risks; therefore, artificial intelligence must evolve from "black box intelligence" to "verifiable intelligence."Heinrich stated that the future of AI lies not in trust, but in verification. 0G Labs is committed to using blockchain technology as a "layer of truth" to ensure that AI possesses auditability, provability, and outcome orientation. It is reported that the report gathers in-depth insights on the future of digital assets from over 60 industry leaders, including Heinrich.

Ethereum founder Vitalik commented on "Musk stated that the new X platform algorithm will be open-sourced in 7 days, including how to determine recommended search content for users, and this process will repeat every four weeks": "If done properly, this will be a very good move. I hope it can be verified and replicated. While it won't solve all problems, it can effectively address the concerns about algorithm transparency raised by myself and many other public members.""Furthermore, I actually think four weeks might be a bit too optimistic, as this means the algorithm needs to be updated frequently to prevent exploitation. The algorithm does need to be verifiable and replicable, so that users who feel they are being shadowbanned, downgraded, etc., can check the execution code of the algorithm to understand why their posts are not being seen."

LazAI Alpha Mainnet Officially LaunchedThe LazAI Alpha mainnet is now officially live, running on enterprise-grade infrastructure, utilizing the QBFT consensus mechanism, and implementing METIS-based settlement on networks based on the Metis SDK (including Hyperion). This marks a significant transition of AI data from a centralized training model to a new stage of verifiable and assetizable on-chain economy.Real-time Anchoring and AssetizationThe Alpha mainnet achieves real-time anchoring and assetization of AI interaction data. All conversations and interactions users have with AI Agents like Lazbubu and SoulTarot will be minted into unique "Data Anchoring Tokens (DAT)." Through the METIS settlement layer and PoS verification mechanism, these can be transformed into on-chain assets with transparent traceability and revenue attribution, becoming verifiable on-chain assets that users can hold, establishing a trustworthy foundation for the AI economy.Developer IncentivesDevelopers can now deploy verifiable AI Agents through LazAI Docs and the Alith framework. Additionally, LazAI will launch a developer incentive program with a total prize pool of 10,000 METIS, covering full-stage support from early prototypes (Ignition Grants) to mature applications (Builder Grants), and providing multi-level ecosystem empowerment, including cross-social channel promotion and user growth funding pools.

Hyperliquid officially clarifies a recent article that made false accusations against it, claiming that Hyperliquid has issues with solvency, integrity, and transparency. The responses to the 10 specific allegations are as follows:System under-collateralized by $362 million: False. The author of the article ignored HyperEVM USDC (parallel to the Arbitrum bridge), with the current total USDC amount being $4.351 billion.Manipulating trading volume through TestnetSetYesterdayUserVlm: False. This is only a testnet feature and cannot be called on the mainnet.Certain users have privileges, such as fee waivers or manipulation affecting airdrops: False. All fees, balances, and transactions are visible on-chain, with no distortion mechanisms.CoreWriter "God mode" can mint, transfer funds, etc.: False. This is a way to send HyperCore operations to HyperEVM smart contracts, with no described privileges.Governance can freeze the chain, with no revocation function: Misunderstanding. Freezing is used for network upgrades, similar to hard forks on other chains. During the POPCAT event in November 2025, L1 was not frozen, only the Arbitrum bridge was automatically locked as a security measure.A single private key can instantly set oracle prices: Misunderstanding. The HIP-3 oracle is configured by the deployer and can use MPC, etc. The perpetual contracts operated by validators use a weighted median price, with no delays to ensure security.Eight undisclosed addresses control all transaction submissions: False. Some transactions have been sent directly by validators, and future upgrades will include MEV and anti-censorship mechanisms.The liquidation cartel has an unfair advantage: Misunderstanding. Only HLP can back liquidations, and deposits are permissionless, with most liquidations processed through the order book.Hidden lending protocols involving over $1 million in funds: False. Portfolio margin, lending, and HLP are publicly announced pre-alpha versions, with documentation on file.ModifyNonCirculatingSupply can change token supply: False. The HIP-1 token supply is fixed, and this function is only for display purposes, not affecting execution.

Filecoin Foundation, FilOz, and contributors to the Filecoin ecosystem have announced the launch of Filecoin Onchain Cloud, a decentralized cloud platform designed to bring verifiable storage, retrieval, and payment capabilities fully on-chain.Filecoin Onchain Cloud transforms core cloud functionalities into programmable, verifiable services. This means that developers can now run actual workloads in an open, verifiable cloud environment for the first time, where data, logic, and payments coexist on-chain. The platform is built on Filecoin's vast decentralized storage network, providing on-chain guarantees. Every interaction is auditable, every payment is conditional, and every service is composable. Developers can now build applications that run directly on the infrastructure, which can prove the location of data, how it is accessed, and when payments are made. The same guarantees used to secure smart contracts now also apply to the data they utilize.

Vitalik posted on social media, "We need open-source and verifiable autonomous driving systems for cars."

ChainCatcher news, Google's latest quantum processor Willow has achieved experimentally verified quantum acceleration, operating at speeds far exceeding the world's strongest supercomputer, Frontier, marking a significant step towards practical quantum computing.Research shows that Willow improves speed by approximately 13,000 times when executing physical algorithms, completing tasks that would take traditional computers 3.2 years in just 2 hours. Experts point out that although this breakthrough is still in the research phase, it brings quantum computing closer to practical application and poses potential risks to the elliptic curve cryptography that Bitcoin relies on. Christopher Peikert, a professor at the University of Michigan, warns that quantum computing poses "substantial, even existential risks" to Bitcoin in the long term, although short-term threats remain limited.

ChainCatcher message, Hyperliquid co-founder Jeff.hl posted on the X platform stating that all orders, transactions, and settlements of Hyperliquid are executed on-chain, and anyone can verify the settlement process and system solvency without permission. This transparency and neutrality make fully on-chain DeFi an ideal form of global financial infrastructure.He pointed out that some centralized exchanges (CEX) have serious underreporting issues regarding settlement data. For example, if Binance has thousands of settlements in the same second, it only publicly displays one, which could lead to an actual settlement scale being underestimated by a hundred times. Jeff.hl expressed hope that the industry will regard transparency and neutrality as core features of the new financial system.

ChainCatcher news, the Web3 native AI infrastructure protocol LazAI has announced that its testnet is now fully open. This testnet aims to address three core issues in the AI field: data monopoly, lack of quality verification, and unfair profit distribution through programmable Data Anchoring Tokens (DAT), on-chain traceability, and transparent value flow technology.The LazAI testnet aims to create, own, and profit from AI agents and data assets. It breaks down the closed data barriers of AI, ensuring that every output can be trusted, allowing creators to achieve complete ownership and monetization of their earnings in an open-chain AI economy.Highlights of the testnet include: Alith development framework, Data Anchoring Tokens (DAT), and out-of-the-box scalability; developers can now join the testnet to build verifiable AI applications.Users can complete whitelist registration by August 20 to be among the first to mint AI companion DAT assets—Lazbubu—and receive rare rewards.

ChainCatcher message, according to the Web3 asset data platform RootData X tracking data, in the past 7 days, the verifiable computing layer Boundless has been the project with the most new followers among X (Twitter) top personalities. New influential figures following this project include crypto KOL Yuyue (@yuyue_chris), Australian big lion BroLeon (@BroLeonAus), and 2Lambroz (@2lambro).