Hi there,
Wall Street is no longer experimenting with AI — it's deploying it at scale, cutting headcount, and building proprietary foundation models for fraud. This week's post maps every major move happening simultaneously across banking, fintech, and asset management in 2026.
🔥 Featured Post
Wall Street's AI Arms Race: Agentic Finance, Foundation Models for Fraud, and 5,000 Layoffs — All at Once
- BlackRock's Asimov is a live agentic AI platform scanning filings, research notes, and emails to generate portfolio insights for the firm's equity business — COO Rob Goldstein expects it deployed firm-wide by next investor day
- JPMorgan's LLM Suite now reaches 200,000 employees with 450+ AI use cases in production and a $19.8B tech budget for 2026 — Proxy IQ replaced external proxy advisors for 3,000 annual shareholder votes
- Goldman Sachs became the first major bank to deploy Devin (Cognition's autonomous software engineer) to its 12,000-strong developer workforce, with early reports of 3–4x productivity gains
- Feedzai's RiskFM is the industry's first Tabular Foundation Model for financial crime, trained on $9T in annual payments across 120B events — it matches bespoke models on single-customer data and beats them when trained across institutions
- Wall Street cut 5,000 jobs in Q1 despite record profits — and PwC's 2026 AI study found 75% of AI's economic gains going to just 20% of companies
📚 In Case You Missed It
GPT-5.5, Google's 8th-Gen TPU, and Why AI Is Finally Learning to Say 'I'm Not Sure' — GPT-5.5 nearly doubles FrontierMath Tier 4 scores vs. Opus 4.7, Google's TPU 8 superpods hit 9,600 chips and 2 PB memory, and MIT's RLCR slashes hallucination calibration error by 90% — three stories shaping how fast AI moves and how much you can trust it.
Physical AI Hits the Real World: Sony's Ace Beats the Pros, ChatGPT Walks Into the Clinic, and Enterprise Agents Go GA — Sony's Ace robot beats pro table tennis players on the cover of Nature, OpenAI ships ChatGPT for Clinicians plus HealthBench Professional, and Microsoft's Frontier Suite hits GA — physical, medical, and enterprise AI all crossed into real-world deployment this week.
GPT-Image 2, Codex Everywhere, and the Open-Weights Coding Crown Changes Hands — GPT-Image 2 adds native reasoning to image generation, Codex ships 'for (almost) everything,' Z.ai's open-weight GLM-5.1 tops SWE-Bench Pro over GPT-5.4 and Opus 4.6, Meta's Llama 5 lands with 5M-token context, and Oracle inks 2.8 GW with Bloom Energy.
More posts dropping every day. Stay curious.
— Bhanu @ superml.dev