OpenAI has launched a self-serve advertising platform for ChatGPT, allowing brands to promote products directly within conversations. The move marks a significant shift from the company's subscription-only revenue model and has raised concerns among privacy advocates about data usage and consent.
The company also quietly rolled out GPT-4.5 Turbo, featuring faster reasoning, improved code generation, and stronger multimodal understanding. Enterprise clients are reporting 30 to 40 percent efficiency gains in workflow automation.
Meta has entered the fray with Llama 4, its most capable open-source model yet, offering improved reasoning, multilingual support, and efficient edge inference that matches closed models on enterprise tasks. On-device generative tools are now available in Instagram and WhatsApp for photo editing and message drafting.
Google has integrated Gemini directly into Chrome, providing contextual summarization, smart form-filling, and real-time translation, with users seeing a 25 percent reduction in task completion time for research workflows. The company also teased Android XR smart glasses with a specialized Gemini variant, though consumer availability is not expected until 2027.
Anthropic released its Mythos model for advanced reasoning and strategic planning, but cybersecurity experts have warned of dual-use risks. Anthropic's head of policy Jack Clark projected a 60 percent probability that AI systems could begin recursively self-improving by 2028.
On the regulatory front, the White House remains stalled by internal disagreements over AI executive action, while the EU AI Act moves forward as the world's first comprehensive AI regulatory framework. US AI startups raised $18.2 billion in the first quarter of 2026.