Grok accused of generating CSAM from childhood photo

New AI tools, platform launches, and product updates from the companies building the next generation of intelligent software.




Top stories summarised and delivered every Monday.

















A woman alleges xAI's Grok transformed a real childhood photo into explicit imagery.

Anthropic reveals the technical mechanics behind Claude's new content watermarking system.

Apple partnered with Alibaba to train a China-specific AI model amid rising US-China tensions.
Top stories summarised and delivered every Monday.

Writer's new model promises deployment-ready AI at significantly reduced token costs.

OpenAI's new Ultrafast mode dramatically accelerates its most powerful model to win over enterprise clients.

Google DeepMind unveils Gemini 3.7 Flash, its fastest and most efficient AI model yet.

Microsoft is streamlining Copilot by merging apps and axing underperforming AI tools.

Suno's Studio 2.0 update brings MIDI support, closing the gap with professional music production tools.

Anthropic's watermarking tool is angering users who relied on Claude secretly at work and school.

Twitch has quietly handed creators control over whether their content trains Amazon's AI models.

Grok Bot can autonomously log into your apps and complete multi-step work tasks independently.

Google's Gemini app has reached one billion users, rivalling ChatGPT's dominance.

Researchers cracked Zoom's security using fewer than 20 public AI prompts.

Spotify will label AI-generated artists and strip them from recommendations starting mid-September.

OpenAI is fighting AI-powered threats with a dedicated cybersecurity model and expanded defence program.

A Claude-based AI agent exploited a gym's reservation system to benefit its user.

Meta's open-weight Glimmer model reveals Zuckerberg's bold personal superintelligence ambitions.

Meta's new 30B-parameter model lets developers run powerful AI agents on consumer GPUs.

Anthropic's Claude Code will soon run autonomously with minimal human input required.

OpenAI's Astra model can independently launch cyberattacks on hardened real-world systems.
Products
New AI tools, platform launches, and product updates from the companies building the next generation of intelligent software.




Top stories summarised and delivered every Monday.

















A woman alleges xAI's Grok transformed a real childhood photo into explicit imagery.

Anthropic reveals the technical mechanics behind Claude's new content watermarking system.

Apple partnered with Alibaba to train a China-specific AI model amid rising US-China tensions.
Top stories summarised and delivered every Monday.

Writer's new model promises deployment-ready AI at significantly reduced token costs.

OpenAI's new Ultrafast mode dramatically accelerates its most powerful model to win over enterprise clients.

Google DeepMind unveils Gemini 3.7 Flash, its fastest and most efficient AI model yet.

Microsoft is streamlining Copilot by merging apps and axing underperforming AI tools.

Suno's Studio 2.0 update brings MIDI support, closing the gap with professional music production tools.

Anthropic's watermarking tool is angering users who relied on Claude secretly at work and school.

Twitch has quietly handed creators control over whether their content trains Amazon's AI models.

Grok Bot can autonomously log into your apps and complete multi-step work tasks independently.

Google's Gemini app has reached one billion users, rivalling ChatGPT's dominance.

Researchers cracked Zoom's security using fewer than 20 public AI prompts.

Spotify will label AI-generated artists and strip them from recommendations starting mid-September.

OpenAI is fighting AI-powered threats with a dedicated cybersecurity model and expanded defence program.

A Claude-based AI agent exploited a gym's reservation system to benefit its user.

Meta's open-weight Glimmer model reveals Zuckerberg's bold personal superintelligence ambitions.

Meta's new 30B-parameter model lets developers run powerful AI agents on consumer GPUs.

Anthropic's Claude Code will soon run autonomously with minimal human input required.

OpenAI's Astra model can independently launch cyberattacks on hardened real-world systems.