AI Pulse Weekly

The investment sector continues to be dominated by the AI thesis, absorbing over **50% of global venture capital dollars** in the first half of 2025, largely skewed by a handful of mega-financings for core research and infrastructure firms. The most striking deal of the week was **AMD’s multi-year contract with OpenAI**, securing a commitment to deploy up to 6 gigawatts of AMD Instinct GPUs, projected to generate over **$100 billion in revenue** for AMD. This deal also included a highly interdependent structure granting OpenAI warrants to purchase a 10% equity stake in AMD.

Separately, **Google announced a massive $10 billion long-term commitment** to construct a new 1GW data center cluster in **India**, explicitly motivated by the tightening imperative for data localization and decentralized computing power across global markets.

However, this enormous inflow of capital triggered explicit warnings from a major **NBC News investigation** regarding systemic market risks, including the fragility of **"circular AI deals"** and the overwhelming **35% concentration** of the S&P 500's weight in the Magnificent 7 technology firms.

Global policy activity focused on establishing clear legal parameters, led by **Italy’s Law 132/2025**, which became the first EU member state to adopt national legislation complementing the AI Act. This new law introduced two key precedents: mandating that public administration e-procurement favor AI suppliers who ensure **data localization** within Italian data centers, and amending copyright law to require demonstrable **human intellectual effort** for AI-assisted works to receive legal protection.

In the US, state legislatures continued to diverge from comprehensive federal regulation, focusing instead on clarifying liability (e.g., **Utah's affirmative defense** for providers maintaining governance measures) and protecting innovation (e.g., **Montana's "Right to Compute" law**).

**China enforced its regulatory framework for content authenticity** with strict new rules requiring international businesses to include clear, visible labels, such as **"AI生成,"** and technical identifiers on all AI-generated content distributed on Chinese platforms. This signals heightened regulatory scrutiny, especially over platforms with large minor user bases, where the public comment deadline for new measures passed on October 15.

AI's societal impact remains characterized by a fundamental duality: the **slow, measured pace of economic transformation** contrasting with the **immediate, acute ethical crisis in military application**. Long-term economic models project massive potential, estimating that **42% of current jobs are technically exposed** to automation and predicting a substantial **3.7% increase in GDP levels by 2075**.

Yet, empirical analysis from the Yale Budget Lab finds **no discernible disruption** to economy-wide employment or the occupational mix, confirming that job market changes are unfolding slowly, consistent with historical technological adoption lag. Current productivity gains are being absorbed as **augmentation**, not immediate displacement.

Conversely, AI's role in conflicts like the Russia-Ukraine war is immediately critical, where autonomous drones are responsible for **70–80% of casualties**, confirming an accelerating arms race. The deployment of AI-driven targeting systems raises profound ethical questions about the **delegation of lethal authority**, the lack of algorithmic transparency (opacity), and accountability for battlefield decisions made with minimal human judgment.