Google Photos adds text prompts to its AI image-to-video tool for users 18+, enabling customized video generation with audio by default, while a new Gmail Photos picker simplifies media sharing.

Google's budget-friendly AI Plus plan is now available in all AI plan markets, including the US at $7.99/month, expanding to 35 new regions. It offers Gemini 3 Pro, AI tools, and 200GB storage, targeting users in emerging markets and competing with ChatGPT Go.

Moonshot AI launches open-source multimodal model Kimi K2.5, excelling in coding and video understanding, with performance rivaling top proprietary models.

Google adds follow-up questions to AI Overviews, upgrades to Gemini 3, and introduces AI-powered inbox search for Gmail.

OpenAI launches Prism, a free AI-enhanced workspace for scientific research, integrating with GPT-5.2 to assist with manuscript revision, argument evaluation, and literature search, aiming to accelerate human-led scientific work.

AI2 releases SERA, an open-source coding agent suite with 32B and 8B parameter models. It solves up to 55% of SWE-Bench tasks, outperforming comparable models. SERA is cost-effective, deployable with minimal code, and accessible for custom development.

WhatsApp launches "Strict Account Settings" for high-risk users, blocking unknown contacts, unsolicited files, and alerting to encryption changes.

EU investigates X over Grok AI's sexual deepfakes, assessing DSA compliance risks and potential fines up to 6% of global revenue.

Microsoft launches Rho-alpha VLA model for robots, enhancing adaptability with tactile sensing and natural language control.

OpenAI's ChatGPT ads cost $60 per 1,000 impressions, triple Meta's rate. Advertisers receive only basic performance data, not detailed user action metrics. The company promises to keep user data confidential and not sell it. Initial ads will target free and low-cost tier users, excluding minors and sensitive topics.

Meta has suspended teenagers' access to its AI chatbot personas globally, citing safety concerns and ongoing development for an improved experience. The move follows regulatory scrutiny and lawsuits over AI's potential harm to minors.

Microsoft launches Maia 200 chip, a major AI inference upgrade with over 100B transistors, challenging Nvidia and rivals like Google TPU and Amazon Trainium.

NVIDIA launches Earth-2 AI suite, revolutionizing weather forecasting with accessible, high-precision models and tools for global use.

Claude now supports interactive apps like Slack and Figma, enabling tasks within the chatbot. Available for paid users, it integrates with Claude Cowork for multi-step automation.

YouTubers sue Snap for allegedly using their videos to train AI features without permission, joining similar lawsuits against other tech giants over AI copyright infringement.

LiveKit, a real-time AI voice/video infrastructure provider, raised $100 million at a $1 billion valuation. The funding round, led by Index Ventures, comes just 10 months after its previous financing. LiveKit powers OpenAI's ChatGPT voice mode and serves clients like xAI, Salesforce, Tesla, 911 operators, and mental health services.

Apple plans to overhaul Siri into an AI chatbot integrated into iPhones and Macs, with typing and voice support, set for a June unveiling and September release.

Adobe collaborates with studios to launch Firefly Foundry, a private generative AI model trained on client-owned IP for secure video and 3D asset creation, targeting enterprise production needs.

OpenAI marketing head Kate Ruch countered claims of the company being "woke AI" by stating that some executives, including a co-founder, support Trump and Republican causes.

Google DeepMind acquires Hume AI's CEO and engineers to enhance Gemini's voice capabilities, as major AI firms increasingly target voice technology and bypass regulations by acquiring startup teams.