Google DeepMind, Google’s flagship AI research lab, launched Veo 2, an advanced AI model designed to generate video content.

It builds on the previous Veo model by supporting video clips longer than two minutes and supports up to 4K resolution.

It reportedly shows significant advancements in motion realism, fluid dynamics, and lighting, but challenges like video coherence remain.

This model competes with OpenAI’s Sora, offering a more refined approach to video generation.

Veo 2 was trained using video-description pairs, but concerns about its data sources and impact on the creative industry persist.

To mitigate risks like deepfake misuse, Google employs filtering techniques and watermarking, though these features are still being reined.