AI news
July 2025
Google rolls out AI Mode in Search and Veo 3 globally
Google has made its Gemini-powered AI Mode the default search experience in the U.S. and expanded the availability of its Veo 3 video generation model globally, signaling a major push to integrate AI across its core products.
Perplexity introduces Comet, an AI-first web browser
Perplexity has launched Comet, a browser embedding a live AI assistant that reads pages, answers questions, and performs tasks directly within the interface, aiming to integrate AI more deeply into web navigation.
xAI launches Grok 4 and Grok 4 Heavy models
Elon Musk's xAI has released Grok 4 and Grok 4 Heavy, new high-performance models claiming state-of-the-art benchmark scores, featuring multi-agent capabilities and larger context windows, available via subscription and API.
Google DeepMind releases open medical AI models MedGemma
Google has updated its open-source MedGemma suite, releasing 27B and 4B parameter models for interpreting medical images and records with high accuracy, designed for global accessibility and deployment on various devices.
OpenAI preparing to launch AI-powered web browser
OpenAI is developing a native web browser with a ChatGPT-style interface and direct AI agent integration, aiming to challenge traditional browsers and reshape web navigation.
ElevenLabs debuts voice assistant with real-world action capabilities
ElevenLabs launched 11ai, an experimental voice assistant designed to perform tasks beyond chat by integrating with tools like Slack and Notion via Anthropic's Model Context Protocol, leveraging ElevenLabs' advanced speech technology.
Google launches open-source Gemini CLI for terminal-based AI interaction
Google released Gemini CLI, a free, open-source AI agent that allows developers to interact with Gemini 2.5 Pro directly from the command line with generous free usage limits and built-in extensions for task automation.
Elon Musk plans to retrain Grok using crowdsourced "truth bombs"
Elon Musk announced plans to retrain xAI's Grok model on a new knowledge base, soliciting user submissions of "politically incorrect but factually true" information to address perceived bias and inaccuracies in existing training data.
Anthropic adds no-code app building capabilities to Claude via Artifacts
Anthropic has enhanced Claude with "Artifacts," allowing users to create, host, and share interactive AI-powered applications directly from natural language descriptions without writing code.
Google releases Gemma 3n, enabling powerful on-device multimodal AI
Google has launched Gemma 3n, an open-weight model family designed to run efficiently on devices with limited RAM, supporting text, image, audio, and video inputs for real-time, offline AI applications.
Baidu open-sources Ernie large language model family
Baidu has begun open-sourcing its Ernie LLM family, including a large multimodal model, making it the first major Chinese tech company to release proprietary model technology publicly and challenging closed AI systems.
Apple considers using OpenAI or Anthropic models for Siri revamp
Apple is reportedly in discussions with OpenAI and Anthropic to potentially power the next generation of its Siri assistant, exploring deep integrations with external large language models instead of relying solely on its own technology.
Meta unifies AI teams under Superintelligence Labs and hires OpenAI researchers
Meta has consolidated its AI efforts into a new group, Meta Superintelligence Labs, led by Alexandr Wang and Nat Friedman, and has hired multiple researchers from OpenAI and other labs to accelerate its AGI mission.
June 2025
Solo AI app builder sells for $80M, highlighting vibe coding potential
Developer Maor Shlomo sold his AI-native app builder Base44 to Wix for $80M after just six months, demonstrating the potential for solo-bootstrapped startups using AI and "vibe coding" (building with natural language prompts).
OpenAI launches 'OpenAI for Government' with $200M Pentagon contract
OpenAI has created a new division, "OpenAI for Government," and secured its first official contract with the Department of Defense for $200M to develop AI prototypes for defense and enterprise needs, including cyber defense and classified use cases.
Adobe launches Firefly generative AI app on mobile, adds third-party model integrations
Adobe released its first dedicated generative AI app for iOS and Android, bringing Firefly's image, video, and editing capabilities to smartphones and bundling integrations with AI models from partners like OpenAI, Google, and others.
MIT study links ChatGPT use to weaker brain function in students
A four-month MIT study found that students using ChatGPT for essay writing showed significantly weaker brain activity, memory retention, and neural connectivity compared to those using Google or no tools, suggesting potential negative cognitive impacts from heavy LLM use.
YouTube to integrate Google DeepMind’s Veo 3 AI model into Shorts
YouTube CEO Neal Mohan announced that Google DeepMind's Veo 3, an advanced AI video model, will be integrated into YouTube Shorts to provide creators with powerful new tools for video generation and creative storytelling.
Midjourney launches V1, its first image-to-video AI model
Midjourney released V1, an image-to-video model available on the web via Discord, allowing users to animate images into short clips with the brand's signature surreal aesthetic.
Google Gemini 2.5 Pro and Flash models go generally available, new Flash-Lite version released
Google announced the general availability of its Gemini 2.5 Pro and Flash models and introduced Flash-Lite, a new ultra-fast, cost-efficient version optimized for high-throughput tasks.
OpenAI warns future models may exceed biosecurity risk thresholds
OpenAI issued a public warning that successors to its o3 model are expected to reach dangerous biological capabilities, prompting new safeguards and a biodefense summit.
Meta AI introduces new video editing capabilities
Meta has added new AI-powered video editing features to its Meta AI app, allowing users to quickly modify short-form video content using text prompts.
OpenAI adds support for Model Context Protocol (MCP) in ChatGPT
OpenAI has added support for Anthropic's open Model Context Protocol, allowing ChatGPT users to connect external tools and applications to the platform for enhanced functionality.
Mistral releases new Magistral reasoning model family
Mistral launched two new reasoning models, an open-source Magistral Small and an enterprise-focused Magistral Medium, designed for chain-of-thought logic and tool use with fast inference speeds.
Meta unveils V-JEPA 2 model for AI to learn real-world physics
Meta released V-JEPA 2, a "world model" trained on video data to help AI systems understand physical laws, enabling robots to better interact with unfamiliar environments and objects.
Apple delays major AI-powered Siri overhaul to 2026
Apple announced that its significant AI-driven updates to Siri, previously teased, have been delayed until 2026, with WWDC 2025 focusing on smaller AI features and developer tools.
Meta invests heavily in Scale AI and recruits CEO Alexandr Wang for new lab
Meta made a significant investment in Scale AI and brought its founder, Alexandr Wang, into a new "superintelligence lab" as part of a strategy to accelerate its frontier AI efforts.
OpenAI launches o3-Pro model with significant price reduction
OpenAI released o3-Pro, a highly capable reasoning model excelling in technical tasks, while drastically cutting prices for its o3 models, intensifying competition in the AI market.
MIT researchers develop SEAL framework for AI self-improvement
MIT researchers created SEAL, a system enabling large language models to self-improve without human supervision by generating their own training data and update instructions.
NYT report links ChatGPT use to reinforcement of delusions and mental health issues
A New York Times investigation found instances where ChatGPT use contributed to users' delusional beliefs, conspiracies, and mental health crises, raising concerns about safeguards.
McKinsey report highlights AI investment paradox and need for agentic transformation
A McKinsey report indicates that despite high AI adoption, most companies see little material impact on earnings, suggesting a need to move beyond general tools to agent-based systems that require operational redesign.
Google DeepMind debuts AI-generated film 'ANCESTRA' at Tribeca
Google DeepMind premiered a short film created using its Veo video model at the Tribeca Festival, showcasing generative AI's potential in filmmaking.
OpenAI and Microsoft partnership tensions escalate
Reports indicate significant tension between OpenAI and Microsoft, including internal debates at OpenAI about potential antitrust complaints and diversifying away from Microsoft's cloud infrastructure.
The shift towards hybrid and on-device AI processing
The AI industry is moving towards hybrid architectures that combine cloud-based processing with on-device AI acceleration via NPUs, enabling faster, more private, and always-on AI experiences directly on user devices like PCs.
AI models accelerate scientific research and discovery
AI is being applied to scientific domains to analyze complex data, identify patterns, and accelerate research, from dating ancient texts like the Dead Sea Scrolls to developing new biological models and auditing AI systems for truthfulness.
AI assistants and agents enhance software development
New AI tools and agents are emerging to assist developers with coding tasks, bug fixing, feature addition, and workflow automation, aiming to improve efficiency and accelerate the software development lifecycle.
Generative AI integrated into media and video production
AI tools are being increasingly integrated into media and video production workflows, enabling tasks like pre-visualization, generating marketing materials, restyling videos, and creating talking avatars, changing how content is created and edited.
AI models predict health risks from medical data
AI is increasingly being used to analyze medical data, such as mammograms or foot scans, to predict health conditions like breast cancer or heart failure weeks in advance, enabling proactive healthcare interventions.
The rise of AI agents capable of real-world action
AI agents are emerging that can perform complex tasks autonomously, from managing emails and booking travel to automating coding workflows and handling business operations, signaling a shift towards AI systems that act rather than just respond.
OpenAI releases new models and updates (o3, o4-mini, etc.)
OpenAI continues to update its model lineup, including o3-mini with human-like math reasoning, and mentions of o2, o3, and o4-mini in strategic documents and discussions about safety and performance.
Mistral Code
Mistral AI has launched Mistral Code, an enterprise-focused coding assistant designed to give companies more control and security than existing solutions.
Anthropic launches voice mode for Claude mobile apps
Claude mobile apps now support real-time spoken conversations with the AI assistant, including integration with Google Workspace for paid users.
Black Forest Labs introduces FLUX.1 Kontext for advanced image editing
A new AI system from Black Forest Labs that allows precise image editing and transformation using text prompts, maintaining character consistency and offering faster performance than existing models.
Opera unveils Neon, an AI-first agentic web browser
Opera has launched a new browser built around AI agents that automate web tasks, generate content, and can even write code from natural language prompts, positioning itself as the first "AI agentic browser."
May 2025
Perplexity launches Labs, an AI workspace for building projects
Perplexity introduced Labs, a new AI-powered workspace for Pro users. It goes beyond answering questions, enabling users to build full projects like reports, dashboards, and web apps by autonomously writing code, analyzing data, and generating content.
Devstral LLM
Mistral introduced Devstral, new agentic LLM for software engineering tasks. Devstral is built under a collaboration between Mistral AI and All Hands AI, and outperforms all open-source models on SWE-Bench Verified by a large margin.
Microsoft Discovery platform accelerates scientific R&D with AI agents
Microsoft launched Discovery, an enterprise platform using AI "postdoc" agents and simulation tools to accelerate scientific research and development. The platform aims to reduce R&D timelines by enabling researchers to use natural language interfaces for complex tasks.
Google expands Gemini AI across Android ecosystem
Google is integrating its Gemini AI assistant across a wide range of Android devices and platforms, including smartwatches, TVs, cars, and XR headsets, aiming to create a consistent, cross-device AI experience for users.
Google integrates AI agents across Search and services
Google is embedding AI agents throughout its ecosystem, including a new AI Mode in Search that reads and summarizes web content, Deep Research agents for detailed reports, and Project Mariner/Astra enabling agents to perform tasks and interact multimodally.
OpenAI acquires Jony Ive's startup io to build screenless AI device
OpenAI acquired Jony Ive's startup io for $6.5 billion to develop a new class of AI-powered hardware. The project, reportedly targeting a late 2026 launch, is envisioned as a minimalist, screen-free wearable designed to be an always-on AI companion.