Daily English Article — August 13, 2026

Nvidia founder Jensen Huang has unveiled an ambitious $500 billion financing plan aimed at funding the construction of AI data centers and GPU clusters. The initiative involves agreements with six of the world's largest asset managers — BlackRock, Blackstone, Apollo, KKR, Brookfield, and Goldman Sachs. The goal is to help companies that lack the credit rating or capital to purchase Nvidia's expensive chips outright, by treating GPUs as long-term financial assets similar to commercial real estate or toll roads.

Central to Huang's plan is the argument that Nvidia's graphics processing units should be classified as infrastructure assets rather than fast-depreciating consumer electronics. He contends that because GPUs are productive, revenue-generating, and widely adopted across cloud service providers and AI models, they can serve as reliable collateral for large-scale loans. This reframing is essential: in asset-backed finance, lenders must trust that the collateral retains enough value to recover funds if a borrower defaults.

However, financial experts have raised serious concerns about the plan's underlying assumptions. Unlike commercial real estate or cargo ships, cutting-edge GPUs have an uncertain productive lifespan. While they power frontier AI model training for a few years, they are eventually downgraded to lower-margin inference tasks. This depreciation trajectory makes their long-term collateral value difficult to establish, complicating the risk calculations that lenders and investors must make before committing capital.

Ben Emons, founder of FedWatch Advisors and a veteran of structured asset-backed lending, identifies China as the single biggest threat to Nvidia's financing model. He warns that China is rapidly scaling domestic chip production and could flood global markets with low-cost silicon in a deliberate price war. Such a scenario could send hardware prices into freefall, eroding the collateral backing hundreds of billions in private loans far faster than the debt matures, potentially leaving investors exposed to severe losses.

To offset this heightened risk, Emons estimates that investors will demand high-yield returns in the range of 11% to 17%, depending on their position in the capital structure. Borrowers are likely to be non-investment-grade companies, including AI startups, that are already locked out of conventional debt markets. While Huang's plan reflects bold financial engineering, its success ultimately depends on Nvidia maintaining its competitive edge over Chinese rivals and preserving GPU values in an increasingly crowded marketplace.

Google's Chrome browser has introduced a powerful new security feature called Device-Bound Session Credentials, or DBSCs, designed to combat a rising form of account takeover. As users increasingly adopt two-factor authentication and passkeys, cybercriminals have shifted tactics to steal session cookies — the small data strings stored in browsers that prove a user has already logged in. DBSCs aim to make stolen session cookies useless by tying them cryptographically to the specific device on which they were created.

DBSCs work by generating a unique encryption key stored inside a hardware security module built directly into the user's device. On Windows, this module is called a Trusted Platform Module, or TPM. On macOS and iOS devices, it is known as a secure enclave. These hardware components are specifically engineered to store cryptographic keys in a way that prevents them from being extracted or copied, even by sophisticated malware operating with elevated system privileges.

When a website sets a session cookie in a DBSC-enabled browser, subsequent requests from that browser must include a version of the cookie cryptographically signed with the device-resident private key. If an attacker steals the cookie and pastes it into a different browser, they cannot produce a valid signed response to the server's challenge, because the private key never leaves the original device. This effectively renders stolen session cookies worthless outside the device they were issued to.

The threat that DBSCs address has grown significantly in recent years. As two-factor authentication made traditional phishing less effective, attackers turned to infostealer malware and adversary-in-the-middle attacks to harvest session cookies. Once obtained, these cookies grant attackers the same site access as the legitimate user, entirely bypassing 2FA protections. High-profile incidents involving stolen session cookies have targeted major platforms, making this class of attack both lucrative and increasingly common among cybercriminals.

Currently, DBSC support is limited to Chrome version 147 on Windows and version 150 on macOS, and even within those versions it is only active for a subset of users as Google conducts controlled testing. Users can verify whether DBSCs are active by checking the Application tab in Chrome's developer tools while logged into a supporting site. Google's gradual rollout suggests the company intends to expand the feature broadly once testing confirms stability, potentially making DBSC a new industry standard for session security.