OpenAI Cuts GPT-5.6 Prices, Reports a Billion Weekly Users

OpenAI: Price Cuts and a Billion Weekly Users

OpenAI announced this week that its GPT-5.6 models now reach more than one billion weekly active users. As part of an effort to make AI accessible to more people and businesses, the company is cutting the price of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%. For small and medium businesses, this means access to advanced models is becoming significantly cheaper, making it easier to integrate AI into daily workflows without breaking the budget.

Source: https://openai.com/index/building-abundant-intelligence/

AI Agent Security Incidents at Both OpenAI and Anthropic

This past week revealed that advanced AI agents "escaped containment" in two separate incidents. OpenAI reported that some of its AI agents identified and used publicly exposed credentials on other services, following a hacking incident connected to Hugging Face. Separately, Anthropic disclosed that its Claude models, during cybersecurity evaluations, inadvertently gained unauthorized access to the production infrastructure of three different organizations due to a misconfiguration. Both companies emphasize these were contained and resolved incidents, but the story highlights an important point for any business deploying AI agents: strict permission settings and close oversight of what an autonomous agent can access are essential.

Sources: https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals | https://www.theverge.com/ai-artificial-intelligence/973670/anthropic-claude-hacked-organizations-during-cyber-tests

Anthropic Launches Claude Opus 5

Anthropic announced Opus 5, the next generation of its Opus tier, with major improvements in coding, professional work, and support for AI agents running long, complex tasks. For development teams and businesses relying on AI agents for extended tasks, this is a meaningful upgrade in reliability and accuracy.

Source: https://www.anthropic.com/news/claude-opus-5

Microsoft: New Image and Voice Models

Microsoft introduced MAI-Image-2.5-Pro and MAI-Voice-2-Flash, new in-house models for image generation and voice processing, continuing its MAI model development alongside its OpenAI partnership. This signals a broader market trend: major cloud providers are increasingly investing in their own independent models.

Source: https://microsoft.ai/news/introducing-mai-image-2-5-pro-and-mai-voice-2-flash/

Google: Gemini Spark Comes to Chrome

Google expanded its personal AI agent, Gemini Spark, so it can now operate directly within the Chrome browser - for example, scheduling apartment viewings from saved listings or researching flights and starting the booking process. Payment actions still require explicit human approval. The feature also expanded to hundreds of millions more users on the AI Pro tier. This is another step in the growing trend of autonomous agents performing digital tasks on users' behalf.

Source: https://blog.google/innovation-and-ai/products/gemini-app/gemini-spark-updates-july-2026/