A developer faced over $11,000 in fraudulent charges after a compromised Firebase service account key triggered massive Gemini AI image-generation activity. Although Google’s automated systems suspended the account for abusive activity consistent with a hijack, the company’s billing department refused to waive the debt. The victim followed all recommended remediation steps, yet remains financially liable for the usage that occurred during the 48-hour breach.

Source

Google is testing a new reCAPTCHA verification method that requires users to perform hand gestures in front of their webcams. The system utilizes machine learning to map 21 specific coordinates on the hand, including finger joints and palm geometry, to ensure a live person is present. This experimental feature is designed to combat automated fraud such as credential stuffing and bulk account creation by implementing liveness detection.

Source

Google has updated Slides to allow the Gemini AI engine to generate complete, multi-slide presentations from a single text prompt. Unlike previous iterations that produced static templates, this update creates native layouts where every element, including text and images, remains fully editable by the user. The system includes a planning phase where the AI asks follow-up questions regarding the target audience and tone before generating a reviewable outline.

Source

Google has expanded its Gemini-powered "Take notes for me" feature in Google Meet to individual Google AI Pro and AI Ultra subscribers. Previously restricted to enterprise Workspace accounts, this tool allows the AI to listen to calls, generate real-time transcripts, and produce structured summaries with action items. The resulting documentation is automatically saved as a Google Doc in the host's Drive and distributed via a recap email once the session concludes.

Source

Google Cloud is repositioning Spanner as a unified data engine designed to eliminate the fragmentation common in generative AI architectures. The platform now integrates relational data, graph relationships, vector search, and full-text search into a single, globally distributed architecture. This consolidation allows AI agents to access structured history and semantic context simultaneously without the need for complex data pipelines.

Source

Google has reportedly restricted Meta's access to its Gemini artificial intelligence models after the social media giant requested more computing capacity than Google could provide. This limitation, which began around March, has disrupted and delayed several of Meta's internal AI development projects. While other Google Cloud customers have faced similar constraints, Meta's exceptionally high demand for inference power made it the most significantly impacted client.

Source

Google is introducing a new "Summarize" feature for PDF documents within the Chrome Beta browser. This integration utilizes the Gemini AI model directly inside the built-in PDF viewer to analyze document content. Users can trigger the process via a dedicated button that activates the Gemini side panel for immediate processing.

Source