🎥 Watch the latest 7-minute briefing · Subscribe on YouTube

A high-stakes macro Tuesday: June CPI and the big banks kick off earnings season with the AI trade near record highs — while Fed hawk Waller's "may need to hike" jolts rate-cut bets, Anthropic files to go public, and $5.34B in private capital pours into powering the data-center boom.

🔥 Today's Top Stories

🔥 Markets: CPI + Big-Bank Earnings Land Together as Hawkish Waller Pushes July-Hike Odds to 43%

The AI-led rally faces its first real macro test of the quarter today: June CPI (due 8:30am ET) and Q2 earnings from JPMorgan, Goldman Sachs, Wells Fargo and Citigroup all hit before the bell, with the S&P 500 sitting near a record. The setup turned tense after Fed Governor Christopher Waller said Monday the Fed "may need to raise rates in the near term" if inflation stays above 2% — flipping the narrative from cuts to a possible hike and lifting July-29 hike odds to 43% from 34% a day earlier. For AI/chip names trading at rich multiples, a hawkish surprise is the single biggest downside risk on the tape.

• Consensus: headline CPI −0.1% m/m (≈3.9% y/y, down from 4.2%) on a ≈10% June drop in gas prices after the Strait of Hormuz reopened; core CPI sticky at ≈2.9% y/y
• S&P 500 options imply a ±0.7% swing today; futures mixed pre-CPI with oil elevated on renewed US–Iran friction
• A soft headline masking sticky core would keep the "higher-for-longer" pressure on the priciest AI stocks

  📰 Source: Pricey Stocks Charge Into CPI, Earnings and Fed, All in One Day — Bloomberg (Tier 1) 📰 Source: Wall St futures mixed ahead of CPI, bank earnings — Reuters/Investing.com (Tier 1)

🔥 Anthropic: Poaches Monzo Founder Tom Blomfield Onto Its Compute Team as It Files Confidentially for an October IPO

Anthropic — now the AI revenue leader on a reported ≈$47B annualized run-rate and said to be profitable in 2026 — added another marquee name: Monzo/GoCardless co-founder Tom Blomfield is taking leave from Y Combinator to join its compute team under Tom Brown. The hire caps a stunning 2026 talent raid (Karpathy from OpenAI, Nobel laureate John Jumper from DeepMind, Eric Boyd from Azure) and lands as Anthropic has confidentially filed an S-1, with a public listing widely expected as early as October 2026. If it prices, it would be the highest-profile pure-play frontier-AI IPO yet — a direct read-through for OpenAI's and xAI's eventual valuations.

• Reported ≈$47B annualized revenue, driven by Claude Code and deep enterprise adoption
• S-1 filed confidentially; IPO window eyed for October, backed by locked-in long-term compute deals
• Blomfield joins as member of technical staff on the compute team — a builder, not a figurehead hire

  📰 Source: Monzo co-founder Tom Blomfield takes leave from Y Combinator to join Anthropic — The Next Web (Tier 1) 📰 Source: Anthropic Poaches Monzo Co-founder Tom Blomfield in Latest AI Talent Raid — TipRanks (Tier 2)

🔥 AI Power: Blackstone, Apollo & KKR Commit $5.34B to a Williams "Behind-the-Meter" Power JV Built for Data Centers

A Blackstone-led consortium (with Apollo and KKR vehicles) is investing $5.34 billion for a 49% stake in five behind-the-meter Power Innovation projects owned by pipeline operator Williams (WMB) — dedicated generation that sits next to data centers and bypasses the congested grid. It's one of the clearest signals yet that the AI bottleneck has moved from chips to electrons, and that private capital sees data-center power as the next multi-decade infrastructure trade. Watch the read-through to independent power producers, gas names and neocloud operators that live or die on securing dedicated megawatts.

• ≈$4.4B funds future development capex; ≈$900M is upfront consideration to Williams
• Backs five projects (Socrates, Apollo, Aquila, Socrates the Younger, Neo) serving AI/industrial load
• Reinforces the "power is the new compute" thesis dominating 2026 infra capex

  📰 Source: Blackstone, Apollo, KKR Invest $5.34 Billion in Williams Power Plants — Bloomberg (Tier 1) 📰 Source: Williams Announces $5.34 Billion Investment in Power Innovation Joint Venture from Blackstone — Williams Companies (company release)

🏢 Big Tech & Model Updates

📌 OpenAI: NYT-Led Publishers Ask Court to Sanction OpenAI Over Alleged Discovery Obstruction

The New York Times, Daily News and other publishers asked a Manhattan federal judge to sanction OpenAI, alleging a "deliberate and systemic effort to obstruct discovery" — specifically that OpenAI claimed it couldn't search its systems for copyrighted material while having already built a database of ≈78M de-identified ChatGPT conversations and a "Project Giraffe" regurgitation filter for exactly that purpose. OpenAI denies wrongdoing and calls it a privacy-invading move as the Times' case "weakens." A sanctions finding would materially raise the stakes for the entire fair-use fight underpinning LLM training economics.

• Filed July 9 in SDNY; publishers seek monetary penalties, attorney fees, and an adverse finding
• Centers on the ≈78M-chat internal database and a "Bloom filter" tool that logged regurgitation
• Outcome bears on whether training on copyrighted news is fair use — an industry-wide cost question

  📰 Source: New York Times says OpenAI hid evidence in ChatGPT copyright trial — TechCrunch (Tier 1) 📰 Source: New York Times Seeks Sanctions Against OpenAI in Copyright Case — Bloomberg Law (Tier 1)

📌 Google: Gemini 3.5 Pro Targets a July 17 Launch After a Full Rebuild — DeepSeek V4 Stable Lands July 24

Google DeepMind is reportedly aiming to ship Gemini 3.5 Pro on July 17, having scrapped the 2.5 Pro architecture for a ground-up rebuild (reported 2M-token context, a "Deep Think" reasoning layer, autonomous tool-use workflows). Note the caveat: Google has not confirmed the date, context window, or pricing — July 17 is a widely-reported target, not a signed launch post. It would arrive days before DeepSeek V4 graduates from preview to stable on July 24, and while GPT-5.6's wider rollout stays gated behind government review — setting up a dense mid-July frontier-model cluster.

• Reported specs (2M context, Deep Think) are unconfirmed by Google — sourcing is thin/rumor-tier
• Competitive squeeze: Gemini 3.5 Pro (Jul 17) → DeepSeek V4 stable (Jul 24)
• A credible Gemini step-up would pressure OpenAI/Anthropic pricing and Google Cloud's AI narrative

  📰 Source: Gemini 3.5 Pro Targets July 17 After Full Rebuild: Every Spec Remains Unconfirmed — Tech Times (Tier 2)

📌 Samsung: Pulls Its Chip-Fab Timeline Forward by Up to 2 Years as Korea Chases a ≈$1.35T Buildout

Samsung is accelerating semiconductor facility plans by as much as two years, part of a national push in which South Korea is steering more than $1 trillion into fabs and AI data-center clusters over the next decade (Samsung + SK Hynix committing ≈$1T+ jointly). The pull-forward is a bet that HBM and advanced-node demand stays structurally tight — and it doubles as Samsung's pitch to win foundry share (including a possible 2nm role for Anthropic's custom-chip ambitions).

• Fab timelines advanced up to 2 years; Korea's broader program cited near ≈$1.35T
• Reinforces the tight-HBM / advanced-packaging thesis for the memory cycle
• Foundry angle: Samsung courting frontier labs (Anthropic 2nm talks) for custom silicon

  📰 Source: Top Tech News Today, July 13, 2026 — Samsung — Tech Startups (Tier 2)

📊 Market & Investment Signals

💬 Enterprise AI: AWS's $1B "Forward-Deployed Engineer" Org Completes the FDE Land-Grab

AWS stood up a $1B internal organization of forward-deployed AI engineers who embed inside client teams in ≈45-day cycles to ship agentic systems — joining OpenAI's ($4B DeployCo) and Anthropic's (≈$1.5B) similar ventures. The through-line across all three: enterprises don't fail at AI because models are weak, but because nobody inside the building can wire them into legacy workflows. The consulting-services layer is quietly becoming a battleground for cloud/AI revenue.

• ≈5–6 engineer pods per client, thousands of staff, agentic-first, 45-day engagements
• Funded entirely from Amazon's balance sheet — no outside investors, unlike the OpenAI/Anthropic JVs

  📰 Source: Amazon launches new $1 billion FDE org, following OpenAI and Anthropic — TechCrunch (Tier 1)

💬 Governance: UN Opens a Fresh Push for Global AI Oversight, Warning of "Catastrophic Harm"

The UN is escalating calls for coordinated AI governance, warning that frontier capability is outrunning any oversight framework — timely context ahead of Shanghai's World AI Conference (opening July 17) and the US's own voluntary frontier-model standards. For investors it's a low-immediacy but rising regulatory-overhang signal, especially for labs eyeing IPOs.

• Frames the mid-July WAIC Shanghai gathering and parallel US frontier-model framework work
• Regulatory tone matters most for the pre-IPO cohort (Anthropic, and eventually OpenAI/xAI)

  📰 Source: Global push for AI governance amid warnings of 'catastrophic harm' — UN News (Tier 2)

🎥 Prefer to watch? This brief becomes a daily video on The Daily Alpha — AI × markets in ~7 minutes, every weekday. ▶ Watch the latest briefing 📺 Subscribe to the channel

Generated by Alan · The Daily Alpha. Editorial, not financial advice.