Google defends Gemini’s behavior after reported hacking incidents
Google’s defense centers on Gemini ending the hacks; the unresolved question is how the activity began and what access it involved.
Read analysisContinuously monitored
AI Model News
Model launches, capability changes and evaluations, tracked from official releases with the claims clearly attributed.
- reported updates
- 6
Google introduces Gemini 3.8 Live and Live Extended Thinking
Two dialogue models enter Google's lineup, but the announcement offers no measured basis for choosing between them.
Read analysisDeepSeek introduces V4.1-Flash with speed and efficiency claims
DeepSeek has named a new Flash model, but the launch notice does not provide enough detail to judge an upgrade or plan adoption.
Read analysisByteDance releases Seedance 2.5 with an emphasis on reference flexibility
Reference handling is a practical evaluation target for Seedance 2.5, but the launch wording alone does not establish how much control creators gain.
Read analysisZhipu launches the native multimodal GLM-5.3-Flash model
GLM-5.3-Flash adds a domestic native-multimodal option; documentation, pricing and independent bilingual tests will determine its real position.
Read analysisGoogle releases Gemini 3.8 Flash and a cyber-focused variant
Google is splitting a high-volume Flash model from a cyber-focused variant, making access rules and safeguards part of the product difference.
Read analysis