What changed
in AI, explained.
OpenAI officially previews GPT-5.6 Sol and confirms a limited rollout after a government request
OpenAI's own feed now previews GPT-5.6 Sol, and TechCrunch reports the company is limiting the rollout after a government request while saying such restrictions should not become the norm. Here is what is now official, what is still narrow, and why a small team should care.
OpenAI reportedly delays wider GPT-5.6 release after a White House safety review request
The Verge and TechCrunch report that OpenAI will hold back the broad release of GPT-5.6 after the Trump administration asked it to, sharing the model with select partners first. Here is what was reported, what is still unconfirmed, and why a small team should care.
OpenAI publishes its own data arguing AI agents are taking on bigger work tasks
On 25 June 2026, OpenAI posted How agents are transforming work, making the case that its Codex agent is being handed longer jobs and used beyond developers. Here is what the piece claims, why every figure in it is OpenAI's own, and what a small team should actually take from it.
OpenAI and Broadcom unveil Jalapeño, an LLM-optimized inference chip
OpenAI and Broadcom announced an inference chip built to run large language models, code-named Jalapeño. Here is what was actually announced, why the cost of running AI models matters, and what a small team should and should not read into it.
Anthropic introduces Claude Tag, an always-on AI teammate for Slack and team channels
Anthropic launched Claude Tag, a version of Claude that lives inside Slack and team channels and is meant to pick up your company's context from the conversations it sees. Here is what changed, why team-context AI matters, and how to try it without leaking sensitive information.
Adobe is putting AI assistants inside Photoshop, Premiere, and Illustrator, with a redesigned Firefly studio
The Verge reports that Adobe has started rolling AI assistants, chatbot-style helpers, into Creative Cloud apps like Photoshop, Premiere, and Illustrator, alongside a redesigned Firefly AI studio that edits and generates designs from one interface. This is about assistants built into the apps you already use, not a brand new model. Here is a calm, beginner-friendly read on what it means.
MCP adds Enterprise-Managed Authorization, central admin control for AI tool access
The Model Context Protocol now has a stable extension that lets an organization manage which AI tools and agents can reach which servers through its identity provider, so employees connect with a single login instead of clicking through a separate OAuth prompt for every app.
Google advances AMIE, its medical AI, from diagnosis toward managing health conditions over time, with a Nature publication
Google has published new research on AMIE, an experimental medical AI, showing it could help with managing a health condition over multiple visits, a step beyond reaching a single diagnosis. The work appears in Nature. This is a research study and evaluation, not a product you can use and not a replacement for a doctor. Here is a calm, beginner-friendly read on what it means and what it does not.
Facebook is testing an 'AI Mode' search that answers using public posts across Meta's platforms
According to reporting, Meta is rolling out an 'AI Mode' in Facebook search that returns AI-written answers drawing on public posts and other public information across its platforms. Reports describe it as a feature rollout, not a finished global launch. Here is a calm, beginner-friendly read on what it means and what to check.
Android 17 starts reaching Pixel phones and watches, with new multitasking tools and expanded Gemini features
Google has begun rolling out Android 17 and Wear OS 7 to Pixel devices, adding multitasking tools, parental controls, security features, smartwatch upgrades, and a Pixel Drop that brings newer AI models to phones. Reports describe a staged rollout, not an instant launch on every Android phone. Here is a calm, beginner-friendly read on what it means and what to check.
OpenAI faces a multi-state attorneys-general investigation over its impact on users
A group of US state attorneys general is investigating OpenAI over how ChatGPT affects users, according to reporting. The inquiry touches advertising, engagement, data handling, and the treatment of minors and seniors. Here is a calm, beginner-friendly read on what is happening and why consumer AI tools are being scrutinized.
Anthropic cuts off Claude Fable 5 and Mythos 5 access after a US government order
Days after launching its strongest model, Anthropic says a US government export-control directive citing national security has forced it to block access to Claude Fable 5 and Mythos 5 for foreign nationals. Reporting also says Amazon raised concerns beforehand.