In partnership with
Anthropic just built an AI so good at hacking that they won't let you use it.
Their new Claude Mythos model can find and exploit software vulnerabilities with scary precision so instead of a public release, they're locking it behind "Project Glasswing," a restricted access program for security researchers only.
US regulators are now circling, asking hard questions about what happens when AI companies create weapons they can't safely share.
But wait, there's more:
Google's Gemini now builds interactive charts and gives you dedicated project notebooks
OpenAI drops a new $100/month tier to feed the Codex coding frenzy
I'm Alex. Welcome to L8R by Innov8.
Let's dive deep ๐ฐ
In today's post:
Anthropic builds AI too dangerous to release then gives it to hackers anyway
Google turns Gemini into your personal data playground
OpenAI squeezes a new $100 tier between Plus and Pro
The AI That Scared The US Government
Anthropic just built an AI so powerful they refuse to let you use it.
The new Claude Mythos model is a master at finding deep software bugs.
This terrifying skill even sparked an emergency meeting between US regulators and big bank CEOs on April 7.
The details:
Anthropic launched Project Glasswing to give only 40 trusted partners like Apple, Google, and JPMorgan Chase access to the model.
Claude Mythos can spot software flaws that humans missed for decades, including a 27-year-old bug in OpenBSD.
US Treasury Secretary Scott Bessent and Fed Chair Jerome Powell held a secret meeting with big bank bosses to discuss the financial threat.
Anthropic is giving away 100 million dollars in free AI credits to help partners fix critical systems before hackers break in.
Why it matters:
We just crossed a huge line in AI history.
AI makers finally admit their tools can break the world.
The government now treats AI cyber threats as a massive risk to the global financial system.
Public releases of top-tier AI are no longer a sure thing.
๐ก L8R's Take:
Anthropic made the exact right call by locking this beast up.
If a bad actor gets their hands on Mythos, it is game over for global security.
Open-source fans will cry about it, but keeping this AI a secret is the only way to stop a digital disaster.
From our Partner
Smart starts here.
You don't have to read everything โ just the right thing. 1440's daily newsletter distills the day's biggest stories from 100+ sources into one quick, 5-minute read. It's the fastest way to stay sharp, sound informed, and actually understand what's happening in the world. Join 4.5 million readers who start their day the smart way.
Google Gemini Just Became a Real Workspace
Google is turning Gemini from a simple chat box into a serious work hub.
They just dropped two massive features: interactive charts and dedicated project spaces called Notebooks.
This means you can finally organize your messy AI chats into real projects.
The details:
Gemini Notebooks let you group chats, upload specific files, and set custom rules for a single project.
You can now ask Gemini to build interactive charts and graphs right inside your chat window.
You do not need to export your data to Google Sheets to play with the numbers anymore.
These new tools are rolling out now, but only for paid Gemini Advanced and Workspace users.
Why it matters:
AI chatbots are growing up. Before this, long projects in Gemini turned into chaotic, never-ending chat threads.
Now, Google is directly fighting Claude Projects and ChatGPT Plus by offering a true workspace.
It saves you from losing your mind when the AI forgets what you were talking about.
OpenAI wants $100 a month for ChatGPT
Power users finally have a middle option for ChatGPT.
OpenAI just dropped a new $100 per month plan for people who code heavily with AI.
It fills the massive gap between the cheap $20 plan and the crazy expensive $200 tier.
The details:
The new $100 tier gives you five times the usage limits of the basic $20 ChatGPT Plus plan.
If you sign up before May 31, 2026, you get a massive 10x usage boost for Codex, OpenAI's coding tool.
OpenAI is also changing the $20 plan to stop heavy users from burning through limits in a single day.
This new pricing perfectly matches rival Anthropic, which also has $20, $100, and $200 plans.
Why it matters:
Coding with AI is exploding right now.
Over 3 million people use OpenAI's Codex every week, and usage grew five times bigger in just three months.
Developers were hitting a wall on the $20 plan but did not want to pay $200.
This gives those heavy users exactly what they need to keep building.
๐ก L8R's Take:
I love this move because the $200 plan was always too steep for normal creators.
But nerfing the $20 plan at the same time is a sneaky way to force people to upgrade.
If you write a lot of code, the $100 tier is a no-brainer..
๐ Quick L8R Summary
Anthropic's Mythos: Anthropic built an AI so good at hacking they won't release it publicly.
Gemini upgrades: Google added interactive charts and project notebooks to Gemini.
ChatGPT Pro tier: OpenAI launched a $100/month plan to handle Codex's explosive growth.
๐ฉ Innathe L8R engane undarunnu ๐?
We read every reply - just reply to this email and let us know how we can improve !
Appo adutha L8R il kanaam bieโฆ๐