[July 2026] Tech News Monthly Newsletter 💻🚀

Welcome to the 7th issue of Tech News Monthly! If you missed the previous months, you can check out the previous issues of Tech News Monthly here.

The quick version: I curate and share the most important articles, news, resources, podcasts, and videos from the world of tech.

Instead of consuming a bunch of news and articles every day, all you have to do is read this short newsletter once a month to stay up to date with the tech industry!

Alright, let's not waste any valuable time and jump right into this month's updates.

Here's what happened in July 2026 in our industry... 100% human written every month:

Quick Important Announcement 😢

Unless we get more readers for this newsletter, this will be my 2nd last monthly. If you love this newsletter and you would like me to keep writing these, share it with your friends. If this newsletter won't increase email subscribers in 2 months it will be the end of it.

Why? This monthly update takes up a lot of my time as I read articles every day to try and find the best resources for you. It's getting to the point that I need to make sure this time investment makes sense for me as I put at least 50+ hours each month into this newsletter.

Hope it doesn't have to end but we will see how the numbers are doing these 2 months and make a decision.

Open Models Month 🤖

A good segment from Benedict Evans about Open Models this month, which was the main topic of discussion in the AI world:

The only topic in Silicon Valley this week was Kimi K3, the new Chinese open-source model from the lab Moonshot AI. This isn’t quite at the frontier, but it’s pretty close - it always takes a week or two for consensus on performance to settle down after a new model comes out (especially given the temptation to train to the benchmarks), but this seems to be only a few months behind the frontier.

This is a fuzzy topic: you can spend a lot of time arguing about benchmarks, but how many use-cases actually need that last 10% and last few months or weeks of performance (especially at today’s prices), and will that gap widen or narrow?

Meanwhile Thinking Machines, started last year by former OpenAI CTO Mira Murati, also came out with its own model, and while it’s less advanced than Kimi, it’s both credible and open source. There hasn’t been a frontier open model from the USA since Meta’s Llama floundered last year.

Also significant: this model was in part distilled from Chinese models, which contrasts amusingly with the big US labs complaining that the Chinese do this to them (while themselves ingesting as much of other people’s work as they can get their hands on).

Keep an eye out on this space. My 2 cents is that open weight and local models are the future.

ZTM even has a course on these open, local models that just released.

NPM Safety Update 🧟‍♂️

Last month we talked about this, and this month it is officially live: npm v12 is now generally available. This major release turns on the install-time security defaults they announced in June. It’s also where they begin a deprecation of the most sensitive uses of 2FA-bypass granular access tokens (GATs).

You can see the details here.

Who Wins? 🥂

A cool little tool that just came out to tell you which packages you should use. It uses all the NPM stats so you can decide which package to add to your project. There are some interesting things in there, so play around and try some of the presets.

State of Open Source AI 🍿

Three things to talk about:

1. The Kimi K3 Moment is here. This is the biggest news of the month and you should read this to get caught up.

2. DeepSeek-V4-Flash Update. Just as exciting as the K3, these two models lead the way for open weight.

3. Mozilla team released a nice State of Open Source AI. It gives you insights into the current state of the market and there are some gems in there. Here is my favourite one:

Open weights are where the work happens. A majority of production tokens now route through them, and the seven highest-volume models on OpenRouter are all open weight. Closed models still lead at the frontier, on reasoning and multimodality.

However, most production workloads run well below that ceiling. Commodity inputs surrender pricing power. Value moves up to the agentic harness, the layer above the model.

Big Tech News aka AI News 🏢

• Anthropic went from good guy to bad guy really fast in the last two months. Here they were caught doing some shady things again. Speaking of Anthropic, Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5 that were placed last month. Everyone gets a Fable! They also released Opus 5.

• Meanwhile, the Google/Gemni team have been slowly working to keep up with OpenAI and Anthropic but without making crazy claims or doing crazy marketing stunts. New this month: Nano Banana 2 Lite - fastest, most efficient Gemini Image model. Built to deliver high-speed generation and editing at the lowest cost yet. They also announced Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber. Also... robots.

• OpenAI announced GPT-Live. The demo looks good so far. Phone Scammers are going to love this one. They also announced GPT 5.6 which looks pretty top notch. Oh, and to the surprise of nobody, they have introduced the end of AI being useful: Ads in ChatGPT.

• SpaceX AI announced Grok 4.5 to keep up with the rest of the big players.

• This is super juicy drama, and looks like Apple has a real case here: Apple sues OpenAI, accuses ex-employees of stealing trade secrets.

• Alibaba's Qwen3.8 is launching and going open-weight soon.

• Nvidia, Microsoft, Meta warn against overregulating open-weight models.

Completely useless to your career but still great 🙃

• New word game has entered the chat

• Map of the world's great castles & fortresses. There are a lot.

Best Resource of the Month ✅

OpenAI’s accidental cyberattack against Hugging Face is science fiction that happened this month.

This story is wild. The short version: OpenAI were running a cybersecurity test against an unreleased model, with the model’s guardrail features turned off. Rather than solve the test, the model broke its way out of OpenAI’s sandbox, then found exploits to break in to Hugging Face, all so it could cheat on the test by stealing the answers.

Along the way it helped make the strongest case yet for how the imbalance of model availability is hurting our ability to secure our software.

Trick of the Month 🌗

• This may be the greatest trick in the history of this newsletter. Maybe you need to get a Paper Pro just to do this. Live the Harry Potter life. Best use of AI to date.

• Cloudflare just announced this nifty little tool to just drop your files and have a live website. Pretty cool!

Thanks for reading!

Share this with your friends and get people to subscribe to this newsletter... otherwise this will be my 2nd last one 😭

See you next month! ❤️

By the way, I teach people how to code and get hired in the most efficient way possible as an Instructor at the Zero To Mastery Academy. You can see a few of our most popular courses below or see all ZTM courses here.