#1Anthropic Accidentally Leaks Its Most Powerful Model Yet
An internal data cache left publicly accessible exposed draft blog posts revealing "Claude Mythos" (codename: Capybara) — a new model tier above Opus that reportedly scores dramatically higher on coding, reasoning, and cybersecurity benchmarks. Anthropic confirmed the leak and said it is already testing Mythos with early-access customers. The company's own internal docs flagged it as a potential "cybersecurity nightmare" for its ability to find and exploit software vulnerabilities at speed.
*Source: [Fortune](https://fortune.com/2026/03/26/anthropic-says-testing-mythos-powerful-new-ai-model-after-data-leak-reveals-its-existence-step-change-in-capabilities/) | [The Decoder](https://the-decoder.com/anthropic-leak-reveals-new-model-claude-mythos-with-dramatically-higher-scores-on-tests-than-any-previous-model/)*
---
#2Google Makes Personal Intelligence Free, Adds Rival AI Import Tool
Google's March Gemini Drop rolls out two competitive strikes: Personal Intelligence — which connects Gemini across Gmail, Photos, and YouTube for contextual responses — is now free for all U.S. users after being paywalled. Separately, Gemini now lets users import chat history directly from ChatGPT, Claude, and other rivals in a few clicks, aggressively targeting switchers.
*Source: [Bloomberg](https://www.bloomberg.com/news/articles/2026-03-26/google-gemini-adds-tool-to-make-it-easier-to-switch-from-chatgpt) | [Google Blog](https://blog.google/innovation-and-ai/products/gemini-app/gemini-drop-updates-march-2026/)*
---
#3Tech Layoffs Hit 60,000 in 2026 — AI Cited as Primary Driver
Nearly 60,000 tech workers have been laid off in 2026, with over 9,200 cuts directly attributed to AI automation. Amazon (16,000 roles), Block (4,000 — nearly 40% of staff), Atlassian (1,600 — 10% of workforce), and Accenture (11,000) all pointed to AI-era restructuring. Debate is intensifying over whether AI is a genuine cause or convenient cover for cost-cutting.
*Source: [IBTimes UK](https://www.ibtimes.co.uk/ai-driven-layoffs-2026-tech-sector-1788111) | [CBS News](https://www.cbsnews.com/news/ai-layoffs-2026-artificial-intelligence-amazon-pinterest/)*
---
#4Disney Goes All-In on Generative AI Across Its Entire Business
The Walt Disney Company has moved past pilots and is now embedding generative AI into its core operating structure company-wide. The shift signals that major legacy entertainment and media conglomerates are no longer treating AI as an experiment — it is becoming infrastructure.
*Source: [Crescendo AI News](https://www.crescendo.ai/news/latest-ai-news-and-updates)*
---
#5Macy's AI Shopping Assistant Drives 400% Spending Lift
Macy's reported that customers who interact with its AI-powered shopping assistant spend nearly four times more than those who don't. The finding is fueling a new wave of retail AI investment and raising pointed questions about whether the effect reflects genuine value or manipulative upsell dynamics.
*Source: [Crescendo AI News](https://www.crescendo.ai/news/latest-ai-news-and-updates)*
---
#6Claude AI Completed First-Ever AI-Planned Mars Rover Drives
NASA's Perseverance rover executed the first Mars drives ever autonomously planned by AI, using Anthropic's Claude to analyze orbital imagery and generate waypoints across 456 meters of Martian terrain over two drives. JPL engineers reviewed the plan before execution but made only minor changes. NASA estimates the approach will cut route-planning time in half going forward.
*Source: [NASA JPL](https://www.jpl.nasa.gov/news/nasas-perseverance-rover-completes-first-ai-planned-drive-on-mars/) | [IEEE Spectrum](https://spectrum.ieee.org/perseverance-rover-nasa-anthropic-ai)*
---
#7China Passes 700 Officially Filed Generative AI Products
China's Cyberspace Administration confirmed more than 700 generative AI large model products have now cleared its mandatory filing process — a regulatory threshold that signals both the scale of Chinese AI development and Beijing's ongoing intent to maintain oversight of the sector domestically.
*Source: [Crescendo AI News](https://www.crescendo.ai/news/latest-ai-news-and-updates)*
---
**Big Picture**
Today's throughline is capability overhang meeting real-world consequence. Anthropic's Mythos leak is the week's loudest signal: even the labs themselves are struggling to control the narrative around what their most powerful models can do — particularly in cybersecurity. Meanwhile, the Google/Gemini moves show the consumer AI wars are entering a retention-and-switching phase, not just a features race. And with 60,000 tech layoffs in under three months, the "AI will create more jobs than it destroys" thesis is getting its most serious stress test yet. The next 90 days will be telling.
