Global AI News Daily — 2026-05-24

Top Stories

Anthropic's Project Glasswing Finds Over 10,000 Critical Vulnerabilities

Anthropic's AI-driven cybersecurity initiative, Project Glasswing — powered by the Claude Mythos AI — has uncovered more than 10,000 high- or critical-severity vulnerabilities in some of the world's most "systemically important" software since the program launched last month. The initiative represents a major use case for AI in proactive security, scanning critical infrastructure-level codebases at a scale no human team could match. The disclosure highlights both the offensive potential of AI in security research and the growing role of AI companies in critical infrastructure protection.

Anthropic Lands in London, Pitches Claude for AI-Powered Coding

Anthropic's top executives traveled to the United Kingdom this week, promoting Claude as a safer alternative for automating software development workflows. The visit comes as AI-powered coding tools have entered the mainstream, with engineers increasingly grappling with what widespread adoption means for the future of their profession. Anthropic used the occasion to differentiate Claude on safety grounds in a market where OpenAI's Copilot and GitHub integrations already dominate.

OpenAI Prepares for Highly Anticipated IPO Filing

OpenAI is preparing to file for an initial public offering in the coming weeks, according to reporting from The New York Times. The company would be one of the most eagerly watched IPO candidates in a year already shaping up to be a major one for Silicon Valley listings. CEO Sam Altman has guided the company through a dramatic restructuring from nonprofit to capped-profit to for-profit, setting the stage for a historic market debut.

Company Watch

Anthropic — London Expansion & Project Glasswing: Anthropic is aggressively expanding its enterprise footprint, combining a high-profile UK roadshow for Claude's coding capabilities with the disclosure of Project Glasswing's milestone 10,000+ critical vulnerability discovery. The dual announcements position Anthropic as both a safe enterprise AI provider and a serious player in AI-driven cybersecurity.

OpenAI — IPO on the Horizon: OpenAI is reportedly weeks away from filing for an IPO, which would make it one of the highest-profile public market debuts in tech history. The company has been building toward this moment through structural changes that allow outside investors to hold equity. The offering is expected to generate enormous investor interest.

Google — AI Race Heating Up: Multiple reports this week confirm that Google's Gemini has leapfrogged ChatGPT in perceived relevance and usefulness, following a sweeping set of announcements at Google I/O 2026. Google's strategy involves blending Gemini across all of its core products while simultaneously fighting off manipulation attempts on its AI search results — a challenge the BBC reported on this week.

Policy & Regulation

European Commission and OECD Collaborate on AI Strategy Monitoring: The European Commission and the OECD announced on May 19 a new collaboration to jointly monitor national AI strategies and AI investment levels across member states. The initiative is part of the EU's broader effort to operationalize the AI Act and track implementation progress globally.

Trump AI Executive Order — Aftermath and Industry Pressure: The Trump administration's decision to pull a planned AI executive order last week continues to reverberate. Last-minute calls from tech industry officials — including former AI czar David Sacks — warned the White House that the proposed voluntary safety vetting system could inhibit development. The episode has drawn attention to the deep influence the tech industry wields over Washington's AI governance approach, with critics noting the absence of a coherent regulatory framework.

Industry Moves

OpenAI IPO Filing Imminent: OpenAI is preparing to file for an IPO in the coming weeks in what is set to be a landmark moment for the AI industry. The company's move toward a for-profit structure has been building over the past year, and a successful public offering would cement its status as the dominant consumer and enterprise AI brand.

Anthropic Project Glasswing Milestone: Anthropic's cybersecurity initiative has crossed a significant threshold, with 10,000+ critical vulnerabilities identified. The program, which went live roughly a month ago, signals Anthropic's ambitions beyond conversational AI and into enterprise security — a fast-growing market.

Google vs. AI Manipulation — Quiet Arms Race: Google is quietly building defenses against attempts to manipulate its AI-powered search results, following a BBC investigation that revealed simple methods for getting AI chatbots to output misinformation. Google and other AI companies are racing to patch these vulnerabilities as adversarial prompt engineering becomes more sophisticated.

What to Watch

1. OpenAI IPO Filing: With OpenAI reportedly weeks away from filing, investors and industry watchers should expect the formal S-1 submission to drop soon. This will be a defining moment for the AI sector's relationship with public markets and will likely set valuation benchmarks for the entire industry.

2. Trump AI Executive Order — Revised Version: Following the abrupt withdrawal of the planned AI executive order, the White House is expected to revisit and revise the policy. Watch for a new draft to emerge that attempts to balance industry concerns with at least some safety guardrails.

3. EU AI Act Implementation Progress: With the European Commission and OECD now collaborating on monitoring AI strategies, the coming weeks should bring new data on how member states are progressing with AI Act compliance — a key indicator of how global AI regulation is evolving.

Quick Reads

"AI Commencement Speeches Meet Student Backlash" — Former Google CEO Eric Schmidt was booed by graduates when he told them AI would let them assemble teams of AI agents; the moment captures growing public ambivalence toward AI hype.

"How Google Is Starting to Win the AI Race" — Despite early stumbles, Google's Gemini has leapfrogged ChatGPT in relevance and usefulness, and is set to become ubiquitous across Google's product suite.

"Federal AI Spending in 2026 — Where Does It Stand?" — The Brookings Institution examines the Trump administration's AI spending priorities, finding that the three-pillar strategy (innovation, infrastructure, diplomacy) is reflected inconsistently across federal contracts.