Centralized data networks owned and/or managed by a single entity have been structurally broken for many years. why? Single point of failure. If one entity (or a small number of entities) has access to the database, there is only one “point” that needs to be compromised to gain full access. This is a serious problem for networks that hold sensitive data such as customer information, government files, and financial records, as well as networks that control infrastructure such as power grids.
Billions of digital records will be stolen in 2024 alone, costing an estimated $10 trillion. Notable breaches include nearly all of AT&T's customer information and call records, half of America's personal health information, 700 million end-user records from companies using Snowflake, and 100 records stored on RockYou24. Contains 100 million unique passwords and the Social Security records of 300 million Americans.
Source: Bureau of Statistics2024
This is not just a private sector issue. Governments and critical national infrastructure also rely on centralized networks. Recent notable breaches include records on 22 million Americans stolen from the U.S. Office of Personnel Management, classified government communications from multiple U.S. federal agencies, personal biometric data of 1.1 billion Indian citizens, and continued Chinese intrusions into multiple U.S. Internet service providers.
Hundreds of billions of dollars are spent on cybersecurity each year, yet data breaches are becoming larger and more frequent. It has become clear that incremental products cannot fix these network vulnerabilities. The infrastructure needs to be completely redesigned.
Source: market.us2024
AI magnifies the problem
Recent advances in generative AI have made it easier to automate routine tasks and increase work productivity. But the most useful and valuable AI applications require context—access to sensitive user health, financial, and personal information. These AI models also require huge amounts of computing power, making them rarely run on consumer devices (computers, mobiles) and instead accessing public cloud networks such as AWS to perform more complex inference requests. need to be processed. Given the serious limitations inherent in centralized networks listed earlier, the inability to securely connect sensitive user data to cloud AI is a major hurdle to adoption.
Even Apple made this point during the announcement of Apple Intelligence earlier this year, calling out the need for larger, more complex models in the cloud and the fact that traditional cloud models are no longer viable. said.
They cite three specific reasons.
Privacy and security verification: Provider claims, such as not logging user data, often lack transparency and enforcement. Service updates or infrastructure troubleshooting can result in sensitive data being inadvertently logged. Lack of transparency in runtime. Providers rarely make their software details public, and users cannot see if the service is running unmodified or detect changes, even with open source tools. you can't. Single Point of Failure: Administrators are highly demanding. This is a maintenance-level access, and there is a risk of accidental data leakage or misuse by attackers targeting these privileged interfaces.
Fortunately, the Web3 cloud platform provides the perfect solution.
Blockchain Orchestrated Confidential Cloud (BOCC)
The BOCC network is similar to AWS, except that it is built on completely confidential hardware and managed by smart contracts. Although still in its early stages, this infrastructure has been in development for years and is finally being rolled out by the Web3 project and Web2 enterprise customers. The best example of this architecture is Super Protocol, an off-chain enterprise-grade cloud platform fully managed by on-chain smart contracts and built on a trustless execution environment (TEE). These are secure hardware enclaves that keep your code and data verifiably private and secure.
Source: Super Protocol
The impact of this technology addresses all of Apple's concerns mentioned above.
Privacy and security verification: Public smart contracts that orchestrate the network allow users to verify whether user data was transferred and used as promised. Workload and program transparency: The network also verifies the work done within the confidential TEE, cryptographically proving the correct hardware and data. , and verified that the software was used and the output was not tampered with. This information is also sent on-chain for everyone to audit. Single point of failure: Network resources (data, software, hardware) can only be accessed by the owner's private key. Therefore, even if one user is compromised, only that user's resources are at risk.
Cloud AI presents a disruptive opportunity for Web3, but BOCC can be applied to any type of centralized data network (power grid, digital voting infrastructure, military IT, etc.) and provides excellent validation without sacrificing performance. We can provide you with the privacy and security you need. Or latency. Our digital infrastructure is more vulnerable than ever, but blockchain orchestration can fix it.
