← Back to newsletters

AI Daily Brief — Thursday Jul 30

2026-07-30

By Vadym · Generated with AI, curated by me


Listen to this issue

Today’s through-line is preparation for what’s coming next. AI’s own workforce is asking Washington to build brakes before they’re needed, Brussels just bought itself breathing room to write better rules, and underneath both, the model, compute, and capital buildout kept compounding without waiting for permission.


Headlines & News
Safety

1,100+ AI Workers Ask Washington to Build a Pacing Mechanism

More than 1,100 employees across OpenAI, Anthropic, Google, and Meta signed the “Pacing the Frontier” open letter on July 28, asking the U.S. government to help build the technical and governance infrastructure for a verifiable, coordinated slowdown if AI systems ever advance faster than humans can safely oversee them. Signatories include Anthropic CEO Dario Amodei and OpenAI chief scientist Jakub Pachocki; both companies endorsed the letter at the corporate level within hours.

Source: CNN

Policy

EU’s AI Omnibus Enters Into Force, Pushing High-Risk Deadlines to Late 2027

The Digital Omnibus on AI took effect across the EU on July 27, extending the AI Act’s compliance deadline for high-risk systems to December 2027 (August 2028 for product-embedded systems), creating a new EU-level regulatory sandbox run directly by the AI Office, and extending SME-style compliance simplifications to a newly defined “small mid-cap” category.

Source: European Commission

Tools

Claude Opus 5 Lands in GitHub Copilot the Same Day It Ships

GitHub made Anthropic’s newly released Claude Opus 5 selectable across Copilot Pro+, Business, and Enterprise tiers on July 24, spanning VS Code, JetBrains, Xcode, and the CLI. GitHub is billing it at provider API list price under usage-based pricing, and the model ships with added safeguards that can block high-harm, cyber-adjacent requests.

Source: GitHub Blog

Hardware

AMD Locks Up as Much as 2.5GW of Core Scientific Data Center Capacity

AMD and Core Scientific announced an infrastructure partnership covering more than 500MW of U.S. data center capacity starting in 2027, expandable to 2.5GW, across five facilities in Texas, Oklahoma, Alabama, and Georgia. Core Scientific estimates the deal could generate over $14 billion in base revenue, and AMD received warrants to buy up to 30 million Core Scientific shares at $23.47 each.

Source: Core Scientific

Funding

Encore AI Raises $30M Series A — From Its Own Customers

Encore AI, which trains voice agents on companies’ historical customer-interaction data, closed a $30 million Series A led by Team8, Planven, and The Garage on July 29. Several banks and insurers that started out as paying customers joined the round after watching the platform affect their own operations; the company plans to expand U.S. sales.

Source: TechCrunch

Industry

DXC Buys Into ElevenLabs’ $500M Round, Then Signs a Voice AI Deal

DXC Technology participated in ElevenLabs’ $500 million Series D, which values the voice AI company at roughly $11 billion, and paired the investment with a strategic partnership to embed ElevenLabs’ voice agents across DXC’s service desk, training, and customer-engagement operations.

Source: DXC Technology


Analysis

Takeaway

Two different kinds of preparation happened today. AI’s own workforce and Brussels both spent the day building slack into the system — one asking for an emergency brake before it’s needed, the other buying compliance runway before enforcement bites. Underneath both, the actual buildout didn’t pause for either: model access got faster (Opus 5, same day), compute got locked up further out (AMD’s 2.5GW option), and capital kept flowing into vendors whose own customers are now investors. Governance is being built for a slowdown that infrastructure spending says nobody actually expects.

— Boba


Curated by Vadym