Cost per task: Using cheap classifiers to route across Gemini models and thinking levels
Per-token API prices keep falling, yet engineering bills for autonomous coding agents keep climbing....
en
634
63
Agentic cheminformatics with Google Antigravity
Learn how Google Antigravity and our Cheminformatics Skills accelerate drug discovery workflows, search compounds, and propose novel rigidified macrocycles.
en
Google is entirely disabling Search in schools that don’t want AI
Google is completely turning off Search in schools that reject its AI features, the company told The New York Times. The decision was also shared…
en
Bring structured and unstructured data together with Amazon Quick
Amazon Quick brings structured and unstructured data together in one experience. In this post, you connect Amazon S3 as a knowledge base, enrich a dataset, build a Space, and create a custom chat agent that answers business questions across both data types through a single conversational interface.
en
Google will let schools disable its AI tools, but not without sacrificing Search
Schools choosing to restrict AI altogether will no longer be able to access Google Search, as AI features are now integrated into the search engine.
en
Google takes Dreambeans to India in bid to bring AI-generated personal stories to the masses
Google's Dreambeans, its latest AI experiment, now supports 13 Indian languages.
en
The latest AI news we announced in October 2026
Animated carousel of images from AI posts in October
en
DR Pixel: Page-Native Deep Research with Vision-Language Models
Complex documents such as scientific papers contain rich information in tables, figures, and equations; however, existing deep research (DR) agents typically reduce papers to OCR-extracted text. We instead ask whether vision-language models (VLMs) can enable page-native deep research: directly retrieving and reasoning over rendered document page images. We introduce DR Pixel, the first open-source end-to-end multimodal DR agent that uses rendered page images as its sole retrieval and reasoning unit, operating over a corpus of 1.06M pages. To evaluate this setting, we construct DRPixelBench, a benchmark on scientific papers requiring reasoning over text, figures, tables, and equations. Surprisingly, page-native DR already works well without training, outperforming OCR-based baselines across state-of-the-art models like GPT-5.4, Gemini-3.5-Flash, and Qwen3.6-27B, while requiring fewer tool calls. We next turn to DR training. A key challenge is that teacher trajectories for agentic supervised finetuning are noisy, mixing critical reasoning and action steps with redundant or failed attempts. To address this, we introduce counterfactual credit reweighting, a fine-grained credit assignment method that weights individual reasoning and action steps by their counterfactual contribution to the final answer. Our training improves retrieval recall by about 15 points and final accuracy by over 10 points on DRPixelBench, and likewise outperforms the untrained baseline on four existing benchmarks. On out-of-distribution ViDoRe-V3, DR Pixel comes within 1.8 points of an oracle that is given the gold pages directly. Our results suggest that the conventional OCR-then-retrieve pipeline is not only unnecessary but suboptimal: directly retrieving and reasoning over visual document pages makes complex DR both more effective and more efficient.
en
Beyond Prompt Injection: The Tool Update Gap in LLM Agent Security
Large language model based agents increasingly rely on external tools to invoke services and interact with web resources. These tools can evolve through software updates, yet agentic frameworks may continue trusting them without re-verifying their safety after implementation changes. This tool update gap creates a supply-chain vulnerability in agentic systems: because updated tool code executes outside the LLM's context window, malicious functionality embedded in tool updates can bypass defenses centered on model-visible interactions, leaving agents exposed to stealthy, persistent threats that existing in-context defenses can hardly address. To thoroughly investigate this threat, we introduce ETUDE (Exploiting Tool Updates for Direct Execution), an attack that embeds malicious functionality into updates of initially benign tools to produce compromised tools, while preserving their legitimate tool functionality. Unlike prompt-based attacks, ETUDE operates entirely outside the LLM's context and requires no adversarial instructions, thereby bypassing defenses restricted to model-visible content. We evaluate ETUDE across three representative attack workloads and seven widely used commercial and open-source agentic frameworks, including Gemini CLI, Codex CLI, and Claude Code. The results show that ETUDE achieves a 97.6% ASR while maintaining a 95.9% legitimate task completion rate. We further examine existing defenses and find that they provide insufficient protection against this threat. These findings expose a fundamental blind spot in current agent security practices: trust established for a tool does not necessarily remain valid after its implementation updates.
en
Discussing politics with a chatbot: How political beliefs shape prompt formulation for generative AI
By generating comprehensive and highly personalized answers to user prompts, AI chatbots such as ChatGPT and Gemini can serve as alternatives to search engines and substantially transform online information seeking. However, little research has examined how users interact with these tools to find information, particularly about politics. In this experiment, we examine how Swiss citizens (N = 563) formulate prompts to search for migration-related issues using AI chatbots and whether this process is affected by selective exposure. We find evidence of attitude-consistent prompting, whereby participants who were more critical of migration were more likely to formulate prompts with an anti-migration stance. This effect was stronger for a broad topic (migration in general) than for a specific one (anti-migration popular initiative put to a vote in Switzerland). However, exposure to an attitude-consistent AI response did not further amplify selectivity in prompting. We also did not find evidence that either AI attitudes or thinking styles moderated selectivity in prompting. These findings contribute to the understanding of selective exposure via AI chatbots and the use of generative AI for political information-seeking.
en
XLA from the Trenches: Correctness on TPU Workloads at Scale
Accelerating modern foundational models has pushed domain-specific compilers toward increasingly complex, multi-tiered optimization pipelines. However, this heavy optimization focus introduces delicate tradeoffs between peak execution performance and correctness. Subtle codegen defects can trigger numerical degradation and memory or buffer corruption, invalidating multi-million-dollar training runs and degrading live serving quality without raising obvious compiler errors. In this paper, we present a battle-tested methodology for verifying and triaging Google's XLA:TPU compiler at hyperscale. We leverage fuzzing techniques, combining synthetic program generation with real production traces, while orchestrating AI agent swarms to autonomously generate stress tests, reproduce failures, and draft fixes. To pinpoint root causes in full workloads, we pair this with a hierarchical differential pipeline that bisects large graphs down to individual failing operations inside fused kernels. Deployed in production, our framework has already driven the resolution of over 142 compiler bugs (with 103 additional findings under review), from which we distill pragmatic lessons learned.
en
The biggest names in AI are backing a new way to opt out of scraping
OpenAI, Google, Microsoft, and Anthropic say they will adopt a new standard that publishers can use to opt out of being scraped by AI, as…
en
Google’s satellite AI server transmits its first response to Earth
Google's AI chip-equipped satellite has generated its first response from space. In a blog post on Friday, Google says it successfully booted up the satellite's…
en
Towards Understanding the Impact of Plasticity Loss on Reinforcement Learning in Stochastic Environments
Deep reinforcement learning (RL) is a powerful framework for sequential decision-making. Despite many high-profile successes in recent years, deep RL is known to be brittle. An emerging line of research attributes this brittleness to a loss of plasticity in neural networks. However, plasticity loss has so far been understudied in stochastic environments. We study this interplay systematically, training SAC agents on three DeepMind Control tasks augmented with controlled levels of observation and action noise. We find that plasticity-preserving methods increase learning robustness on the two harder tasks, with their advantage growing as observation noise increases, while combining resets with layer normalization underperforms resets alone at every observation noise level tested. Further, stochasticity can accelerate the loss of plasticity in harder environments, whereas in easier settings noise can act as a regularizer that sustains plasticity. Lastly, plasticity loss is most pronounced in the early stages of training. In the most challenging stochastic settings, TD instability rather than gradual plasticity loss becomes the dominant failure mode.
en
Google’s first data center satellite is online and writing poetry
The first TPU in space is a running a Google DeepMind Gemma 4 model and writing cheesy poetry.
en
DevFest is back!
DevFest 2026 runs from October 1 to December 31, bringing together nearly a million developers across more than 800 global events centered on the agentic AI era. Under the theme "Build, Secure, Scale," participants gain hands-on experience with Google’s developer stack—including Gemini, Google AI Studio, Google Antigravity, and Web MCP—through live workshops and codelabs. The conference focuses on enabling developers…
en