Claude Users Must Choose Data Retention Options, OpenAI Unveils GPT-Realtime for Enhanced Voice Interactions, and Microsoft Introduces Innovative AI Models

In today’s AI updates, Anthropic prompts Claude users to accept new data retention terms by September 28, with opt-in conversation logs stored for five years to improve model accuracy. OpenAI unveils GPT-Realtime and a Realtime API for dynamic speech agents, enabling tools like SIP calling and image input. Microsoft enters the generative voice arena with MAI-Voice-1 and its massive MAI-1-Preview model. Meanwhile, Aurelian secures $14 million to introduce AI-powered assistants in 911 call centers, aiming to alleviate staffing shortages for non-emergency calls while keeping humans focused on urgent responses.

Table of Contents

• Today’s Headlines
• Anthropic Privacy Update: Claude Users Face Five-Year Data Retention Choice
• OpenAI Launches GPT-Realtime and Realtime API for Production Voice Agents
• Microsoft AI Introduces MAI-Voice-1 and MAI-1-Preview Foundation Model
• AI Voice Assistants Enter 911 Call Centers to Ease Staffing Shortages
• Conclusion

Hello, and welcome to AI Tech News Today! I’m Aurora, and I’m here with my co-host, Isabelle. Today is August 29, 2025, and we have some exciting updates from the world of Artificial Intelligence. Let’s get right into it!

Today’s Headlines

1. Anthropic Privacy Update: Claude Users Face Five-Year Data Retention Choice

• Starting this week, Claude users will receive notifications about updated terms, which must be accepted by September 28, 2025.
• New users will select their preferences during signup.
• Those who opt-in will have data from chat sessions retained for five years.
• This move aims to enhance training and safety improvements for the platform.

2. OpenAI Launches GPT-Realtime and Realtime API for Production Voice Agents

• OpenAI has introduced GPT-Realtime, its most advanced speech-to-speech model.
• Enhancements to the Realtime API include:
  • Image input support
  • Session Initiation Protocol calling
  • Enhanced server support
• These updates offer developers more ways to create conversational agents that integrate real-world contexts and tools into their workflows.

3. Microsoft AI Introduces MAI-Voice-1 and MAI-1-Preview Foundation Model

• Microsoft has launched:
  • MAI-Voice-1: A model designed for producing high-quality, expressive audio.
  • MAI-1-Preview: A large-scale foundation model trained using 15,000 Graphics Processing Units.
• These models represent Microsoft’s first in-house developed approaches to AI, showcasing their commitment to advancing speech technology.

4. AI Voice Assistants Enter 911 Call Centers to Ease Staffing Shortages

• Aurelian, a startup backed by Y Combinator, has secured $14 million to deploy AI voice assistants for non-emergency 911 calls.
• These assistants will manage non-urgent inquiries, allowing human dispatchers to focus on critical emergencies.
• The CEO emphasized that this technology is designed to supplement rather than replace human staff.

Conclusion

That wraps up today’s headlines, folks! Be sure to stay informed about the latest developments in AI. There’s always something exciting happening in this fast-evolving field. We hope you’ll join us again for more updates.

Thanks for listening!