Three stories this week highlight different risks in our tech-dependent world. From companies losing direction with AI to critical security flaws and new regulatory moves.

Companies Under “AI Psychosis”

Mitchell Hashimoto (HashiCorp co-founder) tweeted that he believes “entire companies right now under AI psychosis” — making decisions based on AI hype rather than business fundamentals. He didn’t name names, but the pattern is visible across the industry.

This matters because AI psychosis leads to wasted resources and strategic confusion. Companies rush to implement AI everywhere without asking if it solves actual problems. The result? Expensive infrastructure, confused teams, and solutions looking for problems.

For businesses considering AI: start with your pain points, not the technology. At Artemis Lab, we see this constantly — companies want “AI agents” before defining what tasks actually need automation. The smart move is identifying specific workflows that eat up human time, then building targeted solutions.

Google’s Pixel 10 Gets 0-Click Exploited

Google’s Project Zero team published details of a zero-click exploit chain targeting the Pixel 10. The vulnerability allows remote code execution without any user interaction — attackers just need to send specially crafted data to the device.

Zero-click exploits are the worst kind of security flaw. No user error required, no suspicious links to click. Your device gets compromised just by existing on the network. Google has patched this specific issue, but it demonstrates how complex modern devices create attack surfaces we barely understand.

This connects to a broader infrastructure reality: security isn’t just about your code anymore. It’s about every component in your stack, from mobile devices accessing your systems to the cloud infrastructure running your AI agents. Defense requires assuming compromise at multiple layers.

California Targets Dead Online Games

California’s legislature advanced a bill requiring game companies to either keep online games playable after shutdown or provide refunds to players. The bill cleared a key committee and moves toward a full vote.

This isn’t just about gaming. It’s about digital ownership and service continuity. The principle applies to any online service that customers pay for — from SaaS tools to AI platforms. If you sell access to something digital, what happens when you shut it down?

For businesses building AI agents or cloud services, this regulatory trend matters. California often leads national policy. Planning for service discontinuation isn’t just good practice — it might become legal requirement. Build systems that can export user data and function independently when possible.

The common thread across these stories: technology decisions have consequences beyond the immediate technical scope. AI psychosis wastes money and focus. Security flaws compromise entire ecosystems. Service dependencies create regulatory liability. Smart companies plan for all three.