Daily AI Mail

23/05/2026

Anthropic Says Project Glasswing Found 10,000 High-Risk Vulnerabilities
Anthropic says Project Glasswing and Claude Mythos Preview have found more than 10,000 high- or critical-severity vulnerabilities across critical software. The update shifts the debate from whether AI can find bugs to whether the security ecosystem can verify and patch them fast enough.

Story 👇👇

21/05/2026

OpenAI Brings ChatGPT Into PowerPoint for Editable Deck Creation
OpenAI has launched ChatGPT for PowerPoint in beta, letting users create, revise, summarize, and polish presentation decks directly inside Microsoft's slide app. The add-in keeps output editable in PowerPoint while giving ChatGPT access to deck structure and source material.

Story in the first comment

21/05/2026

OpenAI Gives Codex Appshots, Goal Mode, and Team Plugin Sharing
OpenAI is rolling out a new batch of Codex updates led by Appshots, a Mac feature that lets developers attach app context directly to a Codex thread. The release also expands long-running goal mode, browser annotations, business analytics, and shared plugins for teams.

Link to story in the first comment

21/05/2026

Google Triples Gemini Rate Limits in Antigravity After User Pushback
Google is tripling Gemini model rate limits across paid Antigravity tiers and resetting weekly quotas after early users hit caps faster than expected. The quick adjustment shows how agentic coding tools are being tested as much by usage economics as model performance.



Link to the news story in the first comment

21/05/2026

CapCut Is Bringing Video and Image Editing Into Gemini
CapCut says it is partnering with Gemini to let users edit images and videos directly inside Google's AI app. The integration signals a broader push to turn conversational AI tools into practical creative workspaces.


Link to news story in the first comment

20/05/2026

Figma Brings an AI Assistant Into Its Design Canvas
Figma is adding an AI assistant to its collaborative canvas, letting teams generate designs, edit existing work, and run multiple design agents in parallel as competition in AI creative tools intensifies.


Link to news story in the first comment

20/05/2026

NanoClaw Rejects $20M Offer, Raises $12M Seed
NanoCo, the startup behind NanoClaw, raised a $12 million seed round after turning down an acquisition offer near $20 million. The secure, container-based OpenClaw alternative is now chasing enterprise deployments as AI agent tooling heats up.



Link to news story in the first comment

Source: Image Credits: Eyal Toueg / NanoClaw Via TechCrunch

20/05/2026

Google Targets Enterprise AI Costs With Gemini 3.5 Flash
Google introduced Gemini 3.5 Flash at I/O 2026 as a lower-cost frontier model for enterprises running AI agents. The launch also brings Gemini Omni Flash, Gemini Spark, and an updated Antigravity platform as Google pushes deeper into agentic AI.


Link to news story in the first comment

20/05/2026

Alibaba Pushes AI Independence With M890 Chip and Qwen 3.7-Max
Alibaba introduced its Zhenwu M890 AI accelerator and Qwen 3.7-Max model as part of a broader push to reduce reliance on Nvidia chips. The launch shows how Chinese cloud vendors are building more of the AI stack themselves.


Link to news story in the first comment

11/05/2026

11/05/2026

When you’re ready for fewer rate limits and interruptions, just complete the agent import flow in onboarding or go to Settings → General and press “Import agent setup.”