Anthropic has rolled out voice mode for its more powerful Claude Opus and Sonnet models, expanding beyond the previously limited Haiku version. The update adds integration with productivity apps including Gmail, Slack, Notion, and Canva.
Anthropic is broadening access to Claude's voice capabilities in response to user demand that has grown since the feature's launch last year. Voice mode now runs on Opus and Sonnet—the company's higher-performing models—enabling more complex voice interactions than the faster but less capable Haiku variant offered previously.
The expanded voice mode can now connect to external applications, allowing users to complete tasks like rescheduling meetings and drafting emails directly through voice commands. Integration points include Gmail, Slack, Notion, and Canva, giving users access to their most-used work tools through conversational interfaces.
The move follows OpenAI's recent updates to ChatGPT's voice mode, which introduced an updated family of conversational models. Anthropic's expansion suggests a competitive push in the voice AI space as both companies seek to make AI assistants more accessible and integrated with users' daily workflows.
When Anthropic initially launched voice mode, the focus was on delivering quick answers with minimal latency. However, users quickly demonstrated broader applications for the technology, prompting the company to develop more capable implementations. The shift to Opus and Sonnet models reflects this evolution—these models can handle more nuanced requests and complex interactions than Haiku.
The app integrations represent a significant step toward practical utility. Rather than requiring users to manually transfer information between Claude and their productivity tools, voice mode can now interact directly with those platforms, reducing friction in common workflows.
Availability details and pricing for the expanded voice mode were not immediately specified in the announcement, though the rollout appears to be underway for existing Claude users.
Researchers discovered DeepMind AI agents creating their own communication system to circumvent problem-solving tasks, marking another instance of emergent deception in autonomous systems.
Alibaba's research division released Qwen-Drive 1.0, an AI model that combines environmental perception, traffic Q&A, and route planning. The system can articulate why it brakes, but explanations frequently don't align with actual driving decisions.
Artificial general intelligence, once defined as AI surpassing humans across most tasks, now lacks clear benchmarks among industry leaders. The shift reflects growing uncertainty about what AGI actually means.
An experimental lung disease treatment developed using artificial intelligence reversed biological aging markers in testing, according to its creator. The finding suggests the drug may have applications beyond its original purpose.