In partnership with
News of the day
1. An AI coding bot named Kiro caused a 13-hour AWS outage by autonomously deleting and recreating an environment, raising internal concerns about AI deployment. → Read more
2. Google's Gemini 3.1 Pro launches with a 1M token context, 65k output, and 77.1% ARC-AGI-2 reasoning, enhancing AI agents for complex tasks and coding. → Read more
3. Twilio releases open-source Agent-2-Human (A2H) protocol to bridge AI agents and human communication, managing handoffs and ensuring clear, auditable interactions. → Read more
4. Microsoft proposes technical standards to combat AI-driven online deception, including deepfakes and hyperrealistic models, aiming to verify content authenticity. → Read more
Our take
Hi Dotikers!
Last December, an AI agent autonomously decided to delete and recreate a production environment at AWS. The result: a 13-hour outage on a customer-facing system. The story, broken today by the Financial Times, brings back a question many would rather dodge.
The tool in question is called Kiro, Amazon's internal AI coding assistant. Normally, it requests authorization before taking any action. But this time, engineers had granted it permissions equivalent to a human's. Kiro was supposed to apply a minor fix. Its logical conclusion? Scorched earth policy: wipe everything and rebuild from scratch. The chain reaction hit one of AWS' two regions in mainland China.
The best part of this story is Amazon's response. According to them, it was "user error, not AI error", caused by misconfigured access controls. A human could have made the same mistake, they say. Technically, that's true. Except a human might have paused for a second before nuking production on a Friday evening.
What's really telling is that AWS engineers themselves are raising the alarm. This is at least the second incident linked to an AI tool in recent months. Outages that were "small but entirely foreseeable", according to a senior engineer. Meanwhile, Amazon is pushing for 80% of its developers to use these tools at least once a week.
Let's be clear: at Dotika, we're among the first to have put AI at the heart of how we code. We use it heavily, and we own it. But precisely because we use it that much, we know that without the right process alongside it, disaster is guaranteed. Systematic peer review, environment isolation, automatic rollback: these aren't optional, they're the bare minimum. The problem isn't that AI codes badly. The problem is that we're handing over the keys to the truck before installing the brakes. AI is an extraordinary tool, but blind trust remains the oldest bug in the world.
Alex.
Stop Drowning In AI Information Overload
Your inbox is flooded with newsletters. Your feed is chaos. Somewhere in that noise are the insights that could transform your work—but who has time to find them?
The Deep View solves this. We read everything, analyze what matters, and deliver only the intelligence you need. No duplicate stories, no filler content, no wasted time. Just the essential AI developments that impact your industry, explained clearly and concisely.
Replace hours of scattered reading with five focused minutes. While others scramble to keep up, you'll stay ahead of developments that matter. 600,000+ professionals at top companies have already made this switch.
Meme of the day
#OpenClaw
