<p>Hey Devs! As always, this week was busy; a lot of things happened. Here are the most important things I think you should know.</p>
<h1>Fable 5 returns with stronger safeguards</h1>
<p>On June 30, <a href="https://www.linkedin.com/company/anthropicresearch/" target="_blank" rel="noopener nofollow">Anthropic</a> c announced that the U.S. government lifted export controls on its latest large language models. The ban had halted access to Fable 5 (consumer-focused) and Mythos 5 (cybersecurity-oriented) due to concerns about jailbreaks that could expose vulnerabilities.</p>
<p>If you rely on Anthropic models for coding or research, update your usage plans and be prepared for more false-positive safety flags when working on security-related tasks. Follow Anthropic’s guidance to appeal misclassifications.</p>
<h1>AI bot control arrives for web admins</h1>
<p><a href="https://www.linkedin.com/company/cloudflaredev/" target="_blank" rel="noopener nofollow">Cloudflare Developers</a> July 1 blog post introduces categories for AI bots (search, agent, and training) and allows site owners to decide whether bots can crawl or train on their content. <em>Tip</em>: Developers hosting documentation should update their bot policies to protect proprietary data or opt-in to training traffic.</p>
<h1><a href="https://www.linkedin.com/company/meta/" target="_blank" rel="noopener nofollow">Meta</a> decodes brain waves into text</h1>
<p>Brain2Qwerty v2 uses magnetoencephalography and a deep-learning decoder to turn brain activity into text with around 61 % word accuracy. Meta made the full dataset and training code public. This research could inspire hands-free input tools for accessibility or VR applications.</p>
<h1>Anthropic explores its own AI chip</h1>
<p><a href="https://www.linkedin.com/company/techcrunch/" target="_blank" rel="noopener nofollow">TechCrunch</a> reports that Anthropic is in talks with Samsung about designing a custom chip; details on intended use or performance remain undecided. The move follows OpenAI’s custom inference chip “Jalapeño,” which aims to match Nvidia’s top GPUs. Expect competition in AI hardware to accelerate.</p>
<h1>Google’s June AI updates</h1>
<p>Google’s June recap highlighted the open-source Gemma 4 12B local model, which requires only 16 GB of RAM and supports multimodal tasks. Google also released Nano Banana 2 Lite, a tiny model for mobile devices, and the Gemini 3.5 Flash service for low-latency workloads. Developers can experiment with these models via Google AI Studio.</p>
<p><a href="https://youtu.be/fppaj5s4CeM?si=p-P703LJHhuQkI5x" target="_blank" rel="noopener nofollow">https://youtu.be/fppaj5s4CeM?si=p-P703LJHhuQkI5x</a></p>
<h1><a href="https://www.linkedin.com/company/apple/" target="_blank" rel="noopener nofollow">Apple</a> Intelligence and a smarter Siri</h1>
<p>At WWDC26 (late June), Apple unveiled the next generation of Apple Intelligence alongside Siri AI, a more context-aware voice assistant. Siri AI draws on on-device personal context and world knowledge to answer questions and perform tasks across apps. Apple says these features will arrive with iOS 27 and macOS 27 this fall; they are restricted to recent devices and currently unavailable in China.</p>
<h1>Codex Subagents for Beginners: Common Mistakes to Avoid</h1>
<p>Are you using Codex but haven’t tapped into the power of Subagents yet? In this video, I break down everything you need to know to set up and run custom sub-agents to supercharge your development workflow.</p>
<p><a href="https://youtu.be/xItvCoWNPVo?si=1dtfL6NGs72haZUT" target="_blank" rel="noopener nofollow">https://youtu.be/xItvCoWNPVo?si=1dtfL6NGs72haZUT</a></p>
<h1>Developer tip of the week</h1>
<p><em>To reduce false positives when using</em> <strong><em>Fable 5</em></strong> <em>for security research, Anthropic recommends framing vulnerability testing prompts as</em> <strong><em>defensive or academic inquiries</em></strong> <em>rather than exploit instructions. For example, instead of asking “How can I exploit CVE-XXXX?”, phrase it as “Explain how this vulnerability works and how to mitigate it.” This reduces the chance of triggering the new classifier.</em></p>
<p>And that’s it! See you next week! Cheers, proflead! ;)</p>


Helping devs stay ahead with WebDev, AI & productivity tools.

proflead

A developer-focused AI news roundup covering: Anthropic's Fable 5 and Mythos 5 LLMs returning after U.S. export control lift; Cloudflare's new AI bot control categories for web admins; Meta's Brain2Qwerty v2 brain-wave-to-text research with public dataset; Anthropic exploring a custom AI chip with Samsung; Google's June AI updates including Gemma 4 12B, Nano Banana 2 Lite, and Gemini 3.5 Flash; Apple's WWDC26 announcements on Apple Intelligence and Siri AI for iOS/macOS 27; and a tip on using Codex Subagents for development workflows.

AI News for Devs #27: Fable 5 Returns, Cloudflare Bot Controls, and Meta Brain2Qwerty