Google DeepMind released Nano Banana 2 Lite (gemini-3.1-flash-lite-image), the fastest and most cost-efficient image model in the family.
Why it matters: The article gives the speed, pricing, and access points for both models, and explains their positioning and API limits, helping developers assess whether to replace existing workflows.
Google Research has launched TabFM, a foundation model for tabular data classification and regression that frames tabular prediction as an in-context learning (ICL) problem, producing predictions in a single forward pass without manual training, hyperparameter tuning, or feature engineering.
Why it matters: The article explains TabFM's architecture, synthetic-data training approach, and TabArena benchmark results, letting readers weigh it against traditional tuning workflows.
Google Research announced applying its Multi-Token Prediction (MTP) adaptation to the already deployed Gemini Nano v3 model, rolled out to the Pixel 9 and 10 series.
Google Research's linear elastic caching approach, published at CIDR, models page eviction as a ski rental problem, using a lightweight decision tree model to predict TTLs for cached pages and dynamically resize caches. In Spanner production, it cut memory usage by 15.5% versus fixed-size caches while cache misses rose only 5.5%, lowering TCO by about 5% with only a 0.5% impact on actual I/O costs.
A Google Research paper at COLM 2026 found that enabling chain-of-thought reasoning allows Gemini-2.5 (Flash and Pro) and Qwen3-32B to recall simple single-hop facts they nearly always fail to answer when reasoning is turned off.
Google DeepMind announced that computer use is now a built-in tool in Gemini 3.5 Flash, whereas the capability was previously available only as a standalone Gemini 2.5 computer use model.
Why it matters: The article explains that computer use has moved from a standalone model to being built into Gemini 3.5 Flash, and gives an API entry point and enterprise-grade safeguards, making it easier to evaluate automation scenarios.
Import AI issue 462 rounds up multiple items: research from Oxford, the UK AISI, Stanford, and others, via 4 experiments and 18,978 conversations with 6,923 people, found that AI (strongest being Opus 4.1 and Opus 4.6) consistently surpasses expert humans at textual persuasion, raising 10.8 percentage points more in real donations than professional fundraisers; when AI is restricted to human speed and length, the advantage falls to insignificance.
Chrome released an agent-ready toolkit for developers, and Lighthouse gains a new Agentic browsing audit category from M150, checking three metrics: accessibility tree, CLS visual stability, and WebMCP integration.
Why it matters: The article details the specific checks in Lighthouse's new Agentic browsing audit and Chrome DevTools debugging methods, so developers can directly improve how agent-friendly their sites are.
Chrome officially launched third-party developer tools for Chrome DevTools for agents, letting libraries and frameworks expose runtime internal state, such as component hierarchy and server-side state, to AI agents via an event-based JavaScript API.
Chrome 149-150 brings Immediates (also known as push constants) to WebGPU, allowing frequently changing small data to be injected directly into pass encoders via setImmediates(), along with stricter transient attachment validation.
Google DeepMind is partnering with the UK government to develop an AI planning prototype tool built on Gemini, aiming to help planning officers cut decision time on residential planning applications by 50%. The tool is being developed with local planning authorities such as Barnet, Camden and Dorset, and can consolidate data, identify relevant policies, summarize feedback and draft assessment reports, though officers remain the final decision-makers.
Google released a vectorized dataset based on Farmscapes 2020, converting high-resolution raster maps of over 130,000 km² of England into actionable inventories of hedgerows, stone walls, and tree clusters, helping landowners expand forest habitats without compromising food security.
Google DeepMind released the AI Control Roadmap, adding a system-level security layer on top of traditional alignment, sandboxing, and prompt injection defenses. It treats internal agents as potentially misaligned "insider threats," builds a threat modeling framework based on MITRE ATT&CK, uses trusted AI supervisors to detect and block harmful actions, and measures effectiveness by coverage, recall, and response time.
Google released two dermatology AI studies, including a JAMA Dermatology survey of 2,345 people showing the AI-assisted group identified conditions with 23% accuracy, nearly triple the control search group's 8%, though the ability to judge next steps like seeking care did not significantly improve.
At AISTATS 2026, Google Research introduced Regularized f-Divergence Kernel Tests, a framework for auditing machine unlearning with greater sensitivity and flexibility.
Google DeepMind released DiffusionGemma, an open-source experimental model that generates whole blocks of text via a text diffusion approach, achieving up to 4x inference speedups on dedicated GPUs (over 1,000 tokens/second on an NVIDIA H100 and over 700 tokens/second on an RTX 5090).
Why it matters: The official release specifies the hardware behind the 4x speedup and output quality trade-offs, letting readers judge whether it suits local low-concurrency workloads.
Google DeepMind, together with Schmidt Sciences, the Cooperative AI Foundation, and ARIA, and supported by Google.org, launched a research grant program of up to $10M for multi-agent AI safety, open to researchers worldwide.
Google DeepMind released the audio model Gemini 3.5 Live Translate, offering near-real-time speech-to-speech translation across 70+ languages with automatic language detection. It continuously generates translations that preserve the speaker's tone, rhythm, and pitch, lagging the speaker by only a few seconds.
Why it matters: The official announcement details the model's capabilities, access options, and product integrations, helping readers judge how it changes real-time speech translation workflows.
Google DeepMind released Gemma 4 12B, a unified, encoder-free multimodal model in which vision and audio inputs go directly into the LLM backbone, positioned between E4B and the 26B MoE.
Why it matters: The article details the encoder-free architecture, 16GB-memory runnability, and Apache 2.0 open sourcing, helping readers evaluate on-device deployment options.
Google DeepMind launched the Google DeepMind Accelerator: Robotics, a program for early-stage European robotics startups, selecting 15 companies across Europe in its first cohort.
Chrome announced the WebMCP origin trial in Chrome 149, letting developers build structured tools for web apps that declare the purpose of buttons, input fields, and other UI elements and manage page state, so agents can complete tasks without guessing.
Why it matters: With the official WebMCP origin trial open, developers can let agents operate web forms and debugging flows via declared tools and submit API feedback.
Google DeepMind has released results of a pre-registered trial in Sierra Leone: after eight weeks of using Guided Learning, students' math scores improved by +0.258 standard deviations versus the control group, equivalent to about 1.2 to 1.7 years of typical learning progress; classes where teachers integrated Gemini into about half of lessons, reaching the 12-hour target, progressed even more, about 1.8 to 2.5 years.
Why it matters: Google DeepMind shares teaching details and quantified results from the pre-registered Sierra Leone trial, and has opened a teacher training guide and RCT playbook so readers can learn a replicable research approach.
King's College London, Fudan University, and The Alan Turing Institute released the SocioHack benchmark, containing 72 sandboxed social environments that test LLMs' ability to exploit loopholes in real rule systems after RL training; RL let LLMs rediscover historically patched loopholes with 61.25% recall and 90.85% precision.
Google has launched Cross-Corpus Retrieval, powered by Agentic RAG, on the Gemini Enterprise Agent Platform, now available in public preview.
Google published the PHRM research system in Nature, which uses facial video captured by the front camera after face unlock to estimate heart rate and daily resting heart rate (RHR) in the background via deep learning.
Why it matters: The study covers nearly 700 participants of diverse skin tones and was validated in real-life scenarios, giving readers accuracy data for passive camera-based heart rate measurement plus open data access.
Google Research has open sourced its hydrological modeling framework on GitHub under the Apache 2.0 license, allowing national weather and hydrology agencies to integrate AI flood forecasting into their own workflows.
At I/O 2026, Google Research announced Gemini for Science, a set of experimental scientific tools built in collaboration with Google Cloud, Google DeepMind, and Google Labs.
Google has released a zero-trust security design for private analytics that combines a new cryptographic aggregation protocol with trusted execution environment (TEE) hardware protections, provably ensuring Google only receives anonymized, aggregate group insights.
Google DeepMind launched the first cohort of its Accelerator program in Asia Pacific, themed “AI for the Planet.” The three-month program is open to regional startups, research teams, and nonprofits using frontier AI to tackle environmental problems in nature, climate, agriculture, and energy. Selected organizations receive expert mentoring and support integrating frontier and science AI models; it kicks off with an in-person bootcamp in Singapore and applications are now open.
Google's research tool ERA uses Gemini to write and optimize scientific code, and the related paper was published in Nature today. ERA uses tree search to consider thousands of approaches and reaches expert-level performance on benchmarks in genomics, public health, satellite imagery, neuroscience, time series forecasting, and mathematics.
Why it matters: The article gives ERA's performance on benchmarks across multiple disciplines and eight application cases, letting readers understand how AI-assisted scientific computing is actually being deployed.
The labs of biologists Omar Abudayyeh and Jonathan Gootenberg use Google DeepMind's Co-Scientist to accelerate aging research. After scanning tens of thousands of papers, the system proposed over 20 novel candidate genetic factors, some experimentally verified to drive cells into a younger, better-functioning state. Large-scale screening analysis that once took up to 6 months can now be done in days.
Google DeepMind added a Street View grounding capability to Project Genie, letting Google AI Ultra users generate interactive worlds from real US locations in styles like "Ocean World" and "B&W film", powered by Maps Imagery Grounding.
Google DeepMind launched Gemini Omni Flash, the first model in the Gemini Omni family, which can combine image, audio, video, and text inputs to generate videos, with multi-turn editing via natural language and support for character consistency and physical plausibility.
Why it matters: The official introduction of the Omni family details the capability boundaries of conversational video editing, multi-input composition, and physical realism.
Google DeepMind released Gemini for Science, launching three experimental prototypes in Google Labs, including Hypothesis Generation based on Co-Scientist.
On May 19, Google announced an expansion of its content transparency and verification tools, covering Search, Gemini, Chrome, Pixel, and Cloud.
Why it matters: The original maps out how SynthID watermarking and C2PA Content Credentials verification are being extended across Search, Gemini, Chrome, and Pixel, along with rollout timing, to show where content provenance tools stand.
Google DeepMind announced a new National Partnerships for AI initiative in Singapore, focused on healthcare, science, and education. The collaboration includes exploring an AI co-clinician "trio of care" with public healthcare clusters, using AlphaFold to accelerate outbreak research, developing a Gemma-powered running assistant for visually impaired runners, and providing Gemini for Education to all teachers from primary school through junior college.
University of Cambridge professor Clare Bryant uses Google DeepMind's Co-Scientist to study the molecular switches behind how pathogens crossing species (such as avian influenza to humans) trigger severe illness like sepsis.
Matt Onsum, AI/ML lead, and principal scientist Katherine Labbé at Calico Life Sciences use Google DeepMind's Co-Scientist to connect scattered research findings in aging biology and generate hypotheses worth testing.
University of Edinburgh bioengineer Filippo Menolascina used Co-Scientist to study metabolic dysfunction-associated steatohepatitis (MASH), screening candidate combination therapies from a vast body of literature. Addressing why the drug resmetirom works for only some MASH patients, the system proposed the hypothesis that the NLRP3 inflammasome is a molecular bridge linking inflammation and metabolism — a hypothesis later validated experimentally, pointing toward dual-target therapies.
Google DeepMind's Co-Scientist is connecting the labs of MIT mechanical engineer Ritu Raman and Boston Children's Hospital chemical biologist Ryan Flynn to study ALS together.