DeepMind

Twitter

Facebook

Copy Link

Key Facts

Founded2010

Key PeopleDemis Hassabis

HeadquartersLondon, UK

ProductsAlphaFold, Gemini

Top News

US AI giants launch major London expansions as talent and revenue scramble intensifies

Anthropic, OpenAI, Google, and other US AI companies are racing to expand in London, signing record office space in 2026. The city's deep talent pools and financial networks are driving the boom, but the influx is putting pressure on UK startups competing for the same frontier AI talent and resources.

Business2 sources

US AI giants launch major London expansions as talent and revenue scramble intensifies

Business2 sources

Google's DiffusionGemma AI model generates text 4x faster with parallel processing approach

Google DeepMind released DiffusionGemma, an experimental open-source AI model that generates text in parallel blocks rather than sequentially. The model achieves speeds exceeding 1,000 tokens per second on NVIDIA H100 GPUs and 700+ tokens per second on RTX 5090 hardware. While it trades some output quality for speed, it excels at structured tasks like code infilling and constraint-heavy problems.

Technology10 sources

Google's DiffusionGemma AI model generates text 4x faster with parallel processing approach

Technology10 sources

Illinois passes America's strongest AI safety law requiring independent audits of frontier models

Illinois lawmakers unanimously passed SB 315, requiring major AI firms like OpenAI and Anthropic to submit to third-party safety audits and report critical incidents within 72 hours. Governor JB Pritzker confirmed he will sign the landmark legislation, positioning Illinois at the forefront of AI oversight as federal regulation stalls. The law takes effect January 1, 2027.

Policy5 sources

Illinois passes America's strongest AI safety law requiring independent audits of frontier models

Policy5 sources

DeepMind CEO Demis Hassabis predicts AGI by 2030 as Google unveils its Einstein Test

Google DeepMind CEO Demis Hassabis declared at I/O 2026 that Artificial General Intelligence will arrive around 2030, plus or minus a year. He introduced the Einstein Test as the benchmark: train AI on physics up to 1901, then see if it can replicate Einstein's groundbreaking 1905 discoveries. As Google invests $180 billion in AI infrastructure, Hassabis warns that society needs to prepare for what he calls "10 times the impact of the industrial revolution, and 10 times faster."

Technology5 sources

DeepMind CEO Demis Hassabis predicts AGI by 2030 as Google unveils its Einstein Test

Technology5 sources

Singapore secures major AI partnerships with Google and OpenAI, $234M investment announced

Singapore has signed separate agreements with Google and OpenAI to strengthen its position as a global AI hub. OpenAI will invest over $234 million and establish an AI lab in the city, while Google's partnership focuses on healthcare, workforce development, and scientific research through its DeepMind lab.

Business2 sources

Singapore secures major AI partnerships with Google and OpenAI, $234M investment announced

Business2 sources

Google DeepMind pays up to $90 million for Contextual AI talent in licensing deal

Google DeepMind has agreed to hire over 20 researchers from AI startup Contextual AI and license its technology for $80 million to $90 million. The deal includes Contextual AI's CEO Douwe Kiela and marks Alphabet's latest acqui-hire strategy to secure AI talent without formal acquisitions. The move has intensified antitrust scrutiny as regulators view such deals as attempts to circumvent merger rules.

Business4 sources

Google DeepMind pays up to $90 million for Contextual AI talent in licensing deal

Business4 sources

Google's Co-Scientist AI system generates novel hypotheses to accelerate scientific discovery

Google has introduced Co-Scientist, a multi-agent AI system built on Gemini that helps researchers generate and refine novel scientific hypotheses. Published in Nature, the system has already validated new drug repurposing candidates for acute myeloid leukemia through laboratory experiments. Researchers can now access the tool through Google's Hypothesis Generation platform.

Science2 sources

Google's Co-Scientist AI system generates novel hypotheses to accelerate scientific discovery

Science2 sources

AI agents identify promising drug candidates in hours as Google and FutureHouse debut research systems

Two new AI systems published in Nature demonstrate how teams of AI agents can compress months of scientific research into hours. Google's Co-Scientist and FutureHouse's Robin identified promising drug candidates for cancer and eye disease, though human supervision remains essential to prevent AI hallucinations and validate results.

Science3 sources

AI agents identify promising drug candidates in hours as Google and FutureHouse debut research systems

Science3 sources

OpenAI Adopts Google's SynthID AI Watermarking as Detection Tools Expand to Chrome and Search

OpenAI is integrating Google's SynthID watermarking technology alongside C2PA metadata to make AI-generated content easier to identify. The move comes as Google expands SynthID verification tools to Chrome and Search, potentially reaching billions of users. SynthID has already labeled 100 billion images and 60,000 years of audio, but questions remain about whether these systems can effectively combat deepfakes.

Technology9 sources

OpenAI Adopts Google's SynthID AI Watermarking as Detection Tools Expand to Chrome and Search

Technology9 sources

Google DeepMind connects Project Genie to Street View, creating AI simulations of real places

Google DeepMind announced at I/O that Project Genie now integrates with Street View's 280 billion images across 110 countries. Users can generate interactive virtual environments based on real locations, adjusting weather or styles like underwater or desert themes. The AI world model serves dual purposes: training robots and self-driving cars on rare scenarios, and letting consumers explore reimagined versions of actual streets.

Technology8 sources

Google DeepMind connects Project Genie to Street View, creating AI simulations of real places

Technology8 sources

Google unveils Gemini Omni, a multimodal AI that generates videos from any input at I/O

Google introduced Gemini Omni at its I/O developer conference, marking a significant step toward multimodal AI that can create videos from text, images, and audio. The AI content generation tool includes digital avatar capabilities and SynthID watermarking to combat deepfakes, but raises concerns about AI slop flooding social media platforms.

Technology24 sources

Google unveils Gemini Omni, a multimodal AI that generates videos from any input at I/O

Technology24 sources

Google launches Antigravity 2.0 with AI agents orchestration, but CLI shift sparks controversy

Google rolled out Antigravity 2.0 at Google I/O, transforming its agentic coding app into a comprehensive development suite with multi-agent orchestration, a new CLI tool, and SDK. Powered by Gemini 3.5 Flash, the update enables developers to run multiple AI agents simultaneously and build custom workflows. But the forced migration from the open-source Gemini CLI to the closed-source Antigravity CLI has triggered developer frustration over transparency and usage limits.

Technology7 sources

Google launches Antigravity 2.0 with AI agents orchestration, but CLI shift sparks controversy

Technology7 sources

Google DeepMind unveils Magic Pointer, an AI-powered cursor that understands context and voice

Google DeepMind introduced Magic Pointer, an AI-powered mouse pointer that understands screen context and voice commands. Launching on Googlebooks laptops this fall, the feature lets users interact naturally by pointing and speaking instead of typing detailed prompts. Early demos are already available in Google AI Studio, and the technology will soon integrate with Gemini in Chrome.

Technology7 sources

Google DeepMind unveils Magic Pointer, an AI-powered cursor that understands context and voice

Technology7 sources

Google unveils Googlebook, an AI-powered laptop built on Android to challenge premium market

Google announced Googlebook, a new line of AI-native laptops powered by Gemini AI and built on an Android-based operating system. Launching this fall through partners like Dell, HP, Lenovo, Acer, and Asus, these premium devices feature the Magic Pointer cursor and mark a strategic shift 15 years after Chromebooks debuted.

Technology39 sources

Google unveils Googlebook, an AI-powered laptop built on Android to challenge premium market

Technology39 sources

Isomorphic Labs Raises $2.1 Billion to Scale AI Drug Discovery and Advance Clinical Trials

Alphabet-backed Isomorphic Labs secured $2.1 billion in Series B funding led by Thrive Capital, with participation from MGX, Temasek, and the UK Sovereign AI Fund. The Google DeepMind spin-off will use the investment to scale its AI drug design engine and expects to begin clinical trials by late 2026, marking a shift from its earlier 2025 target.

Startups5 sources

Isomorphic Labs Raises $2.1 Billion to Scale AI Drug Discovery and Advance Clinical Trials

Startups5 sources

Prev12 3 4…7 Next