AI News, Sep 15, 2026: Apple releases revamped Siri powered by Google Gemini

← Back to AI Matters

The Brief

Tuesday, September 15, 2026

Today's updates highlight a wave of consumer and enterprise AI releases alongside growing debates over industry safety and development speed.

  1. 01

    Google DeepMind released Gemini 3.8 Live and Live Extended Thinking, two new audio models priced at $1.38 per hour of voice conversation.

    Why it matters

    Developers can build real-time voice applications at a fraction of the cost of competing models like OpenAI's GPT-Live-1.

  2. 02

    Major firms including Nvidia and Palantir scaled back their use of Anthropic's Fable model after learning the company stores usage logs for 30 days.

    Why it matters

    Businesses handling highly sensitive data may face stricter limitations on which AI tools they can safely adopt.

  3. 03

    Meta announced Meta One, a global subscription bundle that provides increased AI usage and enhanced creative tools across Facebook, Instagram, and WhatsApp.

    Why it matters

    Regular users and creators can pay a monthly fee to get higher usage limits and advanced creation tools across Meta's social apps.

  4. 04

    Agility Robotics launched Digit 5, a humanoid robot designed to work safely next to humans in warehouses and factories.

    Why it matters

    Industrial facilities can deploy humanoid robots directly alongside human workers without needing expensive safety cages.

  5. 05

    Salesforce introduced Koa, a reasoning model built on Nvidia's open-weight Nemotron model and optimized for enterprise customer-facing roles.

    Why it matters

    Businesses can deploy more autonomous agents to handle complex sales, marketing, and customer support tasks without human intervention.

  6. 06

    Apple released its rebuilt Siri assistant using Google's Gemini models to handle complex tasks, though the update is currently unavailable in the European Union.

    Why it matters

    Users get a smarter assistant capable of handling multi-step requests and screen context, though EU users are excluded for now.

  7. 07

    Nvidia CEO Jensen Huang told Donald Trump that the company will not support calls from other tech leaders to slow down the pace of AI development.

    Why it matters

    The stance highlights a growing rift among tech leaders over whether to pause frontier AI development for safety reasons.