Wednesday, April 1, 2026

OpenAI’s $122B Shockwave

OpenAI’s $122B Shockwave

Today’s Overview

Good morning, OpenAI just set a new fundraising record while racing toward a unified AI superapp, and that alone would have been enough to anchor the day. But there is also a messy Claude Code leak, fresh evidence that AI trust is slipping even as usage climbs, and a wave of product pushes from Meta, Google, and Slack. Let's dive in.

Top Stories

OpenAI Raises $122B at $852B Valuation

OpenAI announced a $122B funding round at an $852B valuation, setting a new high-water mark for venture fundraising. Amazon, Nvidia, and SoftBank anchored $110B of the raise, and OpenAI said revenue has reached $2B per month as it moves toward a unified AI superapp. The company is folding ChatGPT, Codex, and its agent tools into one product after winding down Sora.

  • The round was led by major backers, with Amazon, Nvidia, and SoftBank anchoring $110B of the raise.
  • OpenAI said its business has reached $2B per month in revenue and enterprise now makes up more than 40% of that total.
  • The product strategy is shifting toward one unified AI superapp that combines ChatGPT, Codex, and agent tools.

Claude Code Source Code Leaks Publicly

Anthropic accidentally exposed the source code behind Claude Code in a public registry, spilling more than 1,900 files and over 500K lines of code. Developers found unreleased features, feature flags, and internal codenames, while Anthropic said the incident was human error and that no customer data was exposed. The leak still became a reputational headache as a GitHub mirror spread quickly.

  • The exposure included more than 1,900 files and over 500K lines of code.
  • Developers also found 44 feature flags plus three unreleased projects, including persistent memory and deep planning.
  • Anthropic said it was human error, not a breach and that no customer data was exposed.

Meta Ships Prescription-Ready AI Glasses

Meta is rolling out its first prescription-ready AI glasses with two new Ray-Ban Meta designs, plus new frame and lens options across its wearable line. The glasses add on-device features like nutrition tracking, WhatsApp message summaries, handwriting support, and pedestrian navigation. Pre-orders start at $499, with in-store availability on April 14.

  • The new models include Blayzer and Scriber with adjustable fittings for prescription use.
  • Software features include on-device nutrition tracking along with message summaries, handwriting support, and navigation.
  • Pre-orders start at $499 and in-store availability begins April 14.

Research & Analysis

AI Use Rises While Trust Slips

A Quinnipiac University poll found AI usage rising even as trust, optimism, and job anxiety moved in the wrong direction. Research was the most common use case among AI users, but the share expecting AI to shrink opportunities jumped sharply and most respondents said government oversight is lacking. The result is a clear gap between adoption and confidence.

  • AI usage increased by 14% while trust and optimism declined.
  • The most common use case was research at 51% followed by writing, school or work projects, and data analysis.
  • Only 5% said AI reflects their interests and 74% said government is not doing enough to regulate it.

Microsoft Lets Claude Critique Copilot Research

Microsoft added Critique and Council to Copilot Researcher, turning it into a multi-model system for reviewing research outputs. One mode uses Claude to check source quality, completeness, and grounding before a report ships, while the other runs two models side by side and highlights where they agree or diverge. The rollout also lands alongside a broader expansion of Copilot Cowork.

  • Critique uses Claude as a second model to review reports before they ship.
  • Council runs both models side by side and flags agreement, disagreement, and unique findings.
  • The update arrives with a broader rollout of Copilot Cowork in Frontier.

Trending AI Tools

  • AWS AI security agents AI agents that detect and remediate application vulnerabilities.

  • Holo3 Claims New OSWorld Computer Use Benchmark Lead

  • Claude Compliance API Lets admins audit logs, monitor activity, and plug into compliance systems.

  • Veo 3.1 Lite A lower-cost video model built for high-volume generation.

Quick Hits

Join the AI Recap Newsletter

Get the latest AI news, research insights, and practical implementation guides delivered to your inbox daily.

By subscribing, you agree to our Terms of Service and Privacy Policy.