Anthropic-Decart Deal: $6bn AI Video Talks and Retail Impact

What happened

Anthropic is reportedly in talks to acquire Decart, an Israeli startup specialising in real-time generative video AI, in a deal said to be valued at around $6bn. If confirmed, this would rank among the largest moves in Anthropic's history and would mark a significant expansion beyond its core conversational AI business.

Decart's technology focuses on generating and manipulating video in real time, a capability distinct from the text-based large language models that have defined Anthropic's product line to date, including its Claude assistant. Reporting on the talks suggests Anthropic is looking to broaden its technical footprint into generative video, though no final agreement has been announced and terms could still change.

Why it matters

Should the acquisition proceed, it would signal Anthropic's ambition to move beyond text and language into multimodal, real-time generative media — a capability set that opens up applications well beyond chatbots. Real-time generative video has obvious implications for sectors that rely on visual content and live interaction, retail among them, where product visualisation, virtual try-on and dynamic content generation are increasingly part of the digital shopping experience.

For technology and transformation leaders, the deal is a reminder that the boundaries between "conversational AI" and "generative media" vendors are blurring quickly. Organisations planning AI roadmaps around single-modality tools may need to revisit those assumptions as leading model providers acquire capabilities that let them offer combined text, voice and video generation from a single platform.

By the numbers

• $6bn — the reported valuation of Anthropic's talks to acquire Decart.

The Renascence take

The headline here isn't really about Anthropic's balance sheet — it's about what happens when the infrastructure behind everyday AI assistants starts to include real-time video generation as standard. That has quiet but real implications for how retailers and service brands might design digital interactions in the near future.

Most commentary on this deal will focus on valuation and competitive positioning among AI labs. What's more interesting from a service-design lens is the sequencing: conversational AI became trusted and familiar first, and only now is generative video being folded in behind the same interface. That ordering matters, because customers have already built mental models of what "talking to AI" means — calm, transactional, text-or-voice. Layering in real-time video without careful pacing risks breaking that trust rather than enhancing it. Retailers and CX leaders watching this space shouldn't rush to bolt on generative video experiences the moment the technology becomes available; the operators who win will be the ones who test, in small increments, whether video actually reduces customer effort — say, in product demonstration or troubleshooting — rather than simply making the interaction more spectacular.