Platform Profile
DeepMind
DeepMind (Alphabet) conducts cutting-edge AI research with applications in science, games, and real-world systems.
Role
Fundamental research
Stack
Alpha* systems, scientific applications
Status
Active
DeepMind focuses on solving intelligence to advance science.
Related Coverage
- OpenAI and Anthropic Tell UK MPs Voluntary AI Safety Pledges Are Not Enough; Google DeepMind and Meta Decline to Back New Laws
At a Commons committee hearing on Oct. 13, Anthropic said there should be binding rules on AI companies and OpenAI backed a statutory framework with some mandatory elements, starting with incident reporting, WIRED reports. Google DeepMind and Meta did not back new legislation. This was an evidence session; officials described the UK's current frontier-AI safety regime as voluntary.
- Biohub's $1.8B Virtual Biology Stack Mixes New Cash With Prior Pledges
Biohub said its Virtual Biology Initiative has grown into a $1.8 billion commitment of funding, data, compute and measurement tech for open, AI-ready biology data. New cash includes $300 million from Google DeepMind, Isomorphic Labs and Meta plus more than $500 million from DOE over five years; Biohub's April $500 million pledge and more than $500 million in prior NIH investment are part of the same stack, not fresh money. Commercial funders get an embargo before data goes public, per Reuters.
- Gemini 4 Argon Opens as a Fairwind-Gated Cyber Frontier — Guardrails Come Later
Google DeepMind’s Koray Kavukcuoglu named Gemini 4 Argon for long-horizon software engineering, legal/finance knowledge work, and cyber defense. Access starts with trusted Fairwind defenders (without cyber guardrails) and the U.S. government’s voluntary pre-release process; paid API and AI Ultra wait on safeguard iteration. Lab-claimed benches lead DeepSWE and AutomationBench; CWE-bench v1 ties peers at 68%.
- Lehane Says OpenAI Has Been Holding Safety Talks With Anthropic and Google
OpenAI’s Chris Lehane said safety discussions with Anthropic and Google DeepMind had been running for several weeks. He said OpenAI did not see a need for an antitrust waiver; that is the company’s position, not a legal determination.
- Gemini 3.8 Live Claims the Independent Speech-to-Speech Board
Google shipped Gemini 3.8 Live and Live Extended Thinking for production voice agents. Google says Extended Thinking tops Artificial Analysis’ Speech-to-Speech Quality Index at 82.6; API audio is $0.005/min in and $0.018/min out. SynthID on all generated audio.
- WeatherNext 3 Adds Hourly Forecasts and Selected 5 km Fields
Google announced hourly WeatherNext 3 forecasts, with selected surface variables at 5 km resolution, beginning rollout across Search, Gemini and Maps. Google cites Brightband’s evaluation for its leading-accuracy claim.
- Gemini 3.8 Flash Ships as Google Splits Cyber Into a Gated SKU
Google launched Gemini 3.8 Flash at a $0.75/$3.75 intro price through year-end, plus 3.8 Flash Cyber for trusted defenders via the new Fairwind program. Public Flash stays the workhorse; Cyber-tier tools stay gated to vetted defenders.
- Gemini Omni 1.1 Flash Is a Director’s API — Extend, Interpolate, Upscale
Google launched Gemini Omni 1.1 Flash for developers: scene extension to 40 seconds from 10 seconds of prior context, first/last-frame interpolation, 360p drafts, 4K upscale, and three-second video references. Production video SKU of Gemini Omni — not a new Flash language model. Customers cited: Adobe Firefly, Figma Weave, Runway.
- Gemini 3.5 Transcribe Brings Speech Recognition Into the Gemini Family
Google launched Gemini 3.5 Transcribe — streaming and file speech-to-text with 4.0%/2.6% Artificial Analysis WER, 70% faster time-to-final than Chirp 3, 85+ languages, and public-preview APIs, with use in Gboard Rambler. Google cites AA’s results; noisy call-center audio is the open question.
- Gemma’s Billion-Download Badge Is Distribution Infra — Not 1B Unique Users
Google DeepMind said the Gemma open-model family passed one billion downloads and 100,000 community variants over two years. It also launched the Awesome Gemma directory. Downloads are not a count of unique users.
- Google Hides the Sparkle — SynthID Stays When the Visible Mark Goes
Google VP Josh Woodward said Gemini and Flow are rolling out a Media watermark toggle that removes the visible sparkle from Nano Banana, Omni, and Lyria outputs — while invisible SynthID and C2PA credentials remain. The toggle will not launch where law requires a visible mark. Same day as Anthropic’s Claude text-watermark explainer.
- Ryanair Bets Gemini Enterprise on Crew, Weather, and Failover
Ryanair and Google Cloud announced a five-year partnership planning Workspace and Cloud deployment to about 35,000 employees, Gemini Enterprise workflows, and DeepMind models for fleet operations, with multi-cloud resilience supporting its target of 300 million passengers by 2034.
- Pixel 11 Puts Gemini Nano on the Lock Screen — Hardware Distribution for the App Stack
Google’s Made by Google event launched Pixel 11, Pixel 11 Pro, and Pixel 11 Pro XL — on-device Gemini Nano running up to 3.5× faster on the new Tensor G6, Magic Capture, HiLight LED cues, and pre-orders from $899/$1,099/$1,299 ahead of an August 20 retail date. Proactive agents on-device, not only in a chat tab.
- Hassabis Goes Strategic; Dean Leaves — DeepMind’s Control Plane Rewrites
Demis Hassabis steps down as Google DeepMind CEO to become GDM Chair and Alphabet Chief Scientist; Koray Kavukcuoglu takes day-to-day SVP; Jeff Dean leaves after 27 years to found public-benefit Discovery Loop with Sanjay Ghemawat and, per wires, Oriol Vinyals and Quoc Le. Not a product ship — a control-plane rewrite for one of the three frontier labs.
- Gemini Robotics 2: Whole-Body Control Is the Product Line—API Access Is Partial
Google DeepMind launched Gemini Robotics 2 (whole-body), ER 2 (embodied reasoning and multi-robot collab), and an on-device tier—demos on Apptronik Apollo 2. Differentiator is Gemini-scale planning plus multi-robot orchestration; public developer access is still only partial at launch.
- Lyria 3.5 in Flow Music: Tempo Control Over Head-to-Head Benches
Google announced Lyria 3.5 in Flow Music, claiming improvements in musicality, lyrics and vocals plus easier tempo and duration control. Its announcement provides no comparative benchmark.
- AI Employees Ask Washington to Develop International Pacing Tools
Reuters reported more than 1,100 AI-company employees signed Pacing the Frontier on July 28, seeking U.S.-backed international coordination tools. OpenAI and Anthropic voiced support; the statement did not set a binding pause or enforcement design.
- Google Launches Gemini 3.6 Flash and 3.5 Flash-Lite; Pro Remains in Partner Testing
Google introduced Gemini 3.6 Flash at $1.50/$7.50 and 3.5 Flash-Lite at $0.30/$2.50 per million tokens. It planned restricted Flash Cyber access, while 3.5 Pro remained in partner testing. Performance and efficiency comparisons are attributed to Google.
- Gemini 3.5 Pro Months Behind: Coding Shortfall After a Public I/O Promise
Reuters, citing Bloomberg, reported Google is months behind on Gemini 3.5 Pro—particularly coding—after Sundar Pichai pointed to a June path at I/O. Strong contemporaneous reporting, not a Google blog confession with metrics. Mid-July changelogs still show Flash-class GA, not Pro.
- DeepMind–Isomorphic Bioresilience: Defense Productized Across Prevent, Detect, Respond
Google DeepMind and Isomorphic Labs published a joint bioresilience program—prevention, detection, and response—citing 15+ partnerships, work to adapt SynthID for DNA synthesis screening, AlphaEvolve for metagenomic surveillance, and an Isomorphic rapid countermeasure unit. Biosecurity leaves the model-card footnote.
- Stop the AI Race March: Legitimacy Outside the Same Offices Shipping Models
About 200 protesters marched from OpenAI to Anthropic and Google DeepMind in San Francisco, organized by Stop the AI Race, calling for a collective pause on new frontier training. Small vs capital stack; durable signal that safety, labor, housing, and power are merging locally.
- Gemini Omni Flash and Nano Banana 2 Lite: Media Stack, Not Another Text Frontier GA
Same day Anthropic restored Fable and shipped Sonnet 5, Google opened Omni Flash video preview and Nano Banana 2 Lite image GA—cheap high-volume images plus API-native conversational video for builders, not a pure text frontier launch.
- AlphaFold Database Adds Millions of Complexes — Interaction Scale
EMBL-EBI, Google DeepMind, NVIDIA, and Seoul National University added millions of AI-predicted protein complexes to the AlphaFold Database — including 1.7M high-confidence homodimers — from predictions for about 30M complexes. Open structural data for systems biology/drug discovery; predictions ≠ wet-lab structures.