At RAISE Summit in Paris, ElevenLabs CEO Mati Staniszewski detailed a run to a reported $600m in revenue, and how the voice startup plans to beat the AI labs.

Perplexity has built an internal coding tool codenamed "Teammate" that could take on Cursor and Claude Code, Business Insider reports. A launch is unconfirmed.

France's ZML releases ZML/LLMD, a free inference server running open-source AI across Nvidia, AMD, Google, Intel and Apple chips at top speed.

At a Paris conference, Mark Cuban argued vibe-coding tools like Lovable and Replit can survive Anthropic and OpenAI by owning the workflow, not just the code.

London startup Fleek has raised $25m in Series B funding, led by early Vinted backer Burda, to scale the AI it uses to sort, grade and price used clothing.

Cloudflare is piloting with OpenAI to feed real-time network signals into ChatGPT's search, a turn for a firm built on blocking AI crawlers.

OpenAI's deployment arm is acquiring applied-AI firm Northslope, its second buy since May, betting that helping enterprises use AI beats better models.

Customers are 3x more likely to use third-party GenAI like ChatGPT than company chatbots for support, Gartner finds, as AI returns lag the spend.

NATO signs a ~€200M, seven-year deal with Accenture and Leonardo to build a secure cloud backbone, the Protected Business Network, for the Alliance.

Frontier AI models are outrunning the benchmarks built to measure their hacking skills, just as a 1 August US deadline for new testing standards looms.

Nomagic says its vision-language-action model, running in live warehouses, halved robot-caused human interventions. Its new AI lab bets on mastery first.

Beijing calls Anthropic's Claude Code a security back-door as US House committees probe American firms using cheap Chinese AI models, CNBC reports.