Top AI news of 2025: Biggest announcements that defined the year

The year 2025 has proven to be transformative for Artificial Intelligence (AI), marked by rapid advancements and groundbreaking releases that brought this technology into the mainstream. Leading the charge, Google made a significant impact with the launch of the Gemini 3 model, which features substantial enhancements in reasoning capabilities, multimodal understanding, and real-time productivity enhancements throughout the Gemini ecosystem. In a competitive response, OpenAI also unveiled its latest iterations, GPT-5 and GPT-5.2, both designed to deliver quicker responses and improved contextual comprehension, alongside a suite of AI-powered tools. This year not only spotlighted major model updates but also showcased remarkable progress in AI-driven creative applications and on-device intelligence. Google's Gemini 3.0 stands out as the tech giant's most advanced AI model to date. With its sophisticated reasoning and enhanced multimodal capabilities, Gemini 3 integrates seamlessly across platforms like Google Search, Workspace, and Android. Its advanced agentic features empower users to strategize complex tasks and interact with applications more fluidly, serving as a collaborative partner in productivity. September saw the launch of Perplexity's AI-native web browser, Comet, which operates on the Chromium framework but incorporates autonomous agentic AI to execute tasks on behalf of users. Comet excels in navigating websites, filling forms, and managing online transactions, further illustrating the growing trend of AI integration into daily digital interactions. In a bid to rival Comet and Google's AI offerings, OpenAI debuted its AI-enhanced browser named Atlas. This innovative tool allows users to undertake multi-step actions, such as online shopping and extensive research, all while offering personalized memory features that enable easy access to prior searches. OpenAI also made waves with the introduction of Sora 2, its second-generation text-to-video model, which gained significant traction upon release. New functionalities like Cameos, a remix culture, and an AI-centric Feed have been introduced, encouraging users to share AI-generated videos. While currently limited to the U.S. and Canada, a wider rollout is anticipated in the coming year. In December, OpenAI further strengthened its offerings with the rollout of GPT-5.2, building upon the initial GPT-5 launch. This upgraded model enhances human-like reasoning and visual interpretation, positioning it as a strong competitor to Google's Gemini 3. Lastly, Adobe made significant strides with its AI suite announcements at Adobe Max, including the Firefly Image Model 5 for generating AI images and new video models. Their suite also features tools like Generate Soundtrack, Generate Speech, and the Photoshop Express AI Assistant, which can execute edits autonomously on a user's behalf. Overall, 2025 has marked a pivotal year in the evolution of AI, setting the stage for future innovations and applications.