Anyone else experiencing widespread issues across Google AI services?

Google Antigravity
, , ,
1

Hi everyone,

Since yesterday (July 15, 2026), I’ve been experiencing several issues across different Google AI services, and it doesn’t seem to be limited to my account. Multiple coworkers are seeing very similar behavior as well, including both Pro and Free users.

The main issues we’ve noticed are:

  • Gemini 3.5 Flash (High) and Gemini 3.1 Pro (High) appear to behave very differently than usual.
    • The models frequently ignore the previous conversation context and respond almost exclusively to the latest prompt.
    • Code quality has degraded significantly compared to what we’ve been getting for months.
    • They struggle to work with existing codebases that they normally handle without issues.
    • We’ve also seen very simple Python code containing obvious syntax errors, as well as incorrect Next.js examples with basic hydration mistakes.
  • Yesterday, we were also unable to create API keys in AI Studio for some time.
  • Today, Jules is essentially unusable for me.
    • It fails while trying to discover my GitHub repositories.
    • The web interface also doesn’t list any repositories.
    • My GitHub integration is connected correctly, permissions look fine, and nothing appears to be misconfigured on my side.

Because these issues are affecting multiple users, different subscription tiers, and multiple Google AI products, I’m wondering if anyone else has been seeing similar problems over the last couple of days.

Has anyone experienced the same behavior or received any information about an ongoing incident?

Thanks!

2

I conducted a brief test, but the issue could not be reproduced in my environment. I plan to investigate further, so please refer to the results if I am able to reI conducted a brief test, but the issue could not be reproduced in my environment. I plan to investigate further, so please refer to the results if I am able to reproduce the issue.produce the issue.