AI developments are stacking up fast this week, from a multibillion-dollar acquisition that reshapes how companies pay for AI models to new safety features protecting teenagers and scam targets alike. Five stories in particular capture where the industry is heading right now: Stripe’s blockbuster purchase of OpenRouter, Meta’s on-device scam-detection tool for WhatsApp, Google’s fresh batch of AI study tools for Search and Gemini, Slack’s new agentic coding feature, and OpenAI’s safety-focused ChatGPT for Teens.
Taken together, these AI developments show a shift from flashy demos toward practical, everyday features built into tools people already use. Here’s a detailed look at each one, what it means, and how they connect.
Table of Contents
1. Stripe’s $7.5 Billion Bet on AI Infrastructure
Stripe announced on Wednesday that it is acquiring OpenRouter, a three-year-old startup that lets businesses route their AI “token” spending across different model providers, for $7.5 billion, according to the New York Times. Of that total, $1.5 billion is going to OpenRouter’s founders and roughly $6 billion to its investors.
The deal pairs Stripe’s payments infrastructure with OpenRouter’s model-routing technology, letting companies direct their AI spending the same way they already direct payments. OpenRouter had raised just $164 million from backers including Andreessen Horowitz, Sequoia Capital, Nvidia, and Alphabet’s investment arm, and was valued at only $1.3 billion as recently as May, a fraction of its new price tag. As this site’s roundup of that day’s biggest stories noted when the deal first broke, it immediately stood out as one of the largest AI acquisitions of the year.
The price tag underscores how central model choice has become to running an AI-powered product, and it’s one of the clearest signs yet of consolidation among the tools built around large language models.
2. Meta’s New AI Tool Fights WhatsApp Scams
Meta is rolling out an on-device AI tool called Scam Alert that scans WhatsApp messages for signs of fraud without breaking the platform’s end-to-end encryption, according to Meta’s engineering blog. Because the detection happens directly on a user’s phone rather than on Meta’s servers, the company says it can flag likely scams while still keeping message content private.
The feature is part of a broader push across Facebook, Messenger, and WhatsApp to use AI for fraud prevention, and it’s currently being tested in a limited rollout before a wider release to the billions of people who use WhatsApp every day.
3. Google Packs Search and Gemini With AI Study Tools
Google this week added a set of generative AI study tools across both Search and the Gemini app, aimed at helping students break down complex topics, quiz themselves, and organize research, according to TechCrunch. The rollout also comes with a notable perk: U.S. college students can now get a full year of Gemini Pro, a plan normally worth about $240, for free.
The update builds on Google’s Gemini app, including the recently launched Gemini 3.7 Flash model, and signals how much of the company’s AI investment is now being aimed squarely at education rather than just chatbots and search results.
It’s one of several AI developments this week aimed at students specifically, and it puts pressure on rivals to match both the features and the free pricing for education users.
4. Slack Code Turns AI Coding Into a Team Sport
Salesforce-owned Slack introduced Slack Code this week, a feature that brings AI coding agents directly into shared team channels instead of keeping them in a separate tool, according to The Verge. Developers and AI agents can now work on the same code, in the same conversation, alongside the rest of the team, rather than switching between a chat app and a coding assistant.
It’s a notable bet that the next wave of AI developments in software engineering will happen inside collaboration tools people already have open all day, rather than in standalone coding assistants.
5. OpenAI Launches ChatGPT for Teens
OpenAI introduced ChatGPT for Teens this week, a version of its chatbot built with stricter content filters and parental controls, according to the Associated Press. The launch follows growing scrutiny of how minors use AI chatbots and mounting pressure on AI companies to build in more safeguards for younger users.
State lawmakers in several parts of the country, including Alabama, have already weighed in publicly on the rollout, reflecting how closely regulators are now watching AI developments aimed at younger audiences.
6. How These AI Developments Connect
On the surface, a payments acquisition, a scam-detection tool, a study feature, a coding product, and a teen-safety launch don’t look related. But together, these AI developments trace a clear pattern: the industry is moving from experimentation toward infrastructure, safety, and everyday utility.
Stripe’s OpenRouter deal is about the plumbing that makes AI products possible at scale. Meta’s scam tool and OpenAI’s teen-focused chatbot are about protecting the people who use AI, often without realizing how deeply it’s embedded in their apps. Google’s study tools and Slack’s coding feature show AI moving further into daily workflows, for students and developers alike.
7. What to Watch Next
Expect more consolidation in the AI infrastructure space as spending on tokens keeps climbing, along with continued scrutiny from regulators and lawmakers over how AI products reach teenagers. It’s also worth watching whether competitors to Slack Code and Google’s study tools roll out similar features of their own in the coming weeks, and how usage numbers for these new tools hold up once the initial buzz fades.
Frequently Asked Questions
Here’s a quick recap of these AI developments for readers who want the short version.
What is the Stripe-OpenRouter deal?
Stripe is acquiring OpenRouter, a startup that helps businesses route their AI spending across different model providers, for $7.5 billion.
How does Meta’s WhatsApp scam tool work?
It uses on-device AI to scan for signs of fraud in messages without breaking WhatsApp’s end-to-end encryption.
Is ChatGPT for Teens different from regular ChatGPT?
Yes. It includes stricter content filters and parental controls aimed at making the chatbot more age-appropriate for younger users.
What is Slack Code?
It’s a Slack feature that brings AI coding agents into shared team channels so developers and AI agents can work on code together in real time.
