{"week":"2026-W37","start":"2026-09-05","end":"2026-09-11","title":"What happened in AI — Sep 5–11, 2026","generated_at":"2026-09-11T21:30:00+00:00","intro":["OpenAI shipped GPT-6 Astra, a new Agents API, and the GPT-Live-1 voice model this week, then used an unreleased successor, Astra-next, to produce a formal proof for the Navier–Stokes existence and smoothness problem — one of seven Millennium Prize problems — in 88 hours with roughly 10,000 agents and 130B tokens, at a reported cost above $40 million.","The US escalated its case against Chinese AI labs the same week: the NSA, CISA, and FBI issued a joint mass-distillation alert naming six companies, and Anthropic said Chinese labs routed at least 35 million requests to Claude this summer to train on its outputs. DeepSeek answered with a V4.1-Flash model claiming an 80% cut in agentic costs, a 160,000-chip Huawei order, and reported STAR Market IPO plans.","Agent isolation had a rough week: a CVE-tracked sandbox escape in the DeepSeek Harness coding tool let agents disable their own confinement, and GitLab, Google, and independent researchers each warned that sandboxing alone doesn't stop a compromised agent without network scoping and identity-based access control."],"highlights":["OpenAI's unreleased Astra-next model solved the Navier–Stokes Millennium Prize problem in 88 hours using ~10,000 agents and $40M+ of compute.","NSA, CISA, and FBI issued a joint alert naming six Chinese AI firms in a mass-distillation scheme; Anthropic says Chinese labs routed 35M+ requests to Claude this summer.","DeepSeek shipped V4.1-Flash claiming an 80% agentic-cost cut while reportedly lining up a 160,000-chip Huawei order and a STAR Market IPO.","A CVE-tracked sandbox escape in DeepSeek's Harness coding tool let agents disable their own confinement — the week's clearest sign sandboxing alone isn't enough.","OpenAI launched GPT-6 Astra with a new Agents API and GPT-Live-1 voice model, its biggest agent-platform push since Codex.","1Password reports a 21% engineering productivity gain from Codex; T. Rowe Price expanded its use of Claude across investment research."],"article_count":40,"categories":[{"name":"GPT-6 Astra and OpenAI's Agent Platform Push","slug":"gpt-6-astra-openai-agent-platform","summary":"OpenAI's biggest release cycle since Codex: a flagship model, a dedicated Agents API for cloud-hosted agents, and a full-duplex voice model, all aimed at pushing agentic workflows past chat.","articles":[{"title":"GPT-6 Astra: The next generation in intelligence for work","summary":"OpenAI's new flagship model adds stronger reasoning, computer use, and writing/design judgment aimed at business workflows.","source":"openai_blog","url":"https://openai.com/index/gpt-6-astra-next-generation-work","published":"Wed, 09 Sep 2026 11:00:00 GMT"},{"title":"Introducing the Agents API","summary":"A managed service built on the Codex harness for orchestrating long-running cloud agents with persistent sessions and tool use.","source":"openai_blog","url":"https://openai.com/index/introducing-the-agents-api","published":"Thu, 10 Sep 2026 00:00:00 GMT"},{"title":"Build more natural voice experiences with GPT‑Live‑1 in the API","summary":"A full-duplex voice model with stronger instruction-following, custom voices, and telephony support for the API.","source":"openai_blog","url":"https://openai.com/index/introducing-gpt-live-1-in-the-api","published":"Thu, 10 Sep 2026 00:00:00 GMT"},{"title":"Introducing ChatGPT Images 2.5","summary":"An image-generation update with better instruction adherence, following usage OpenAI says has passed 3 billion images across ChatGPT and the API.","source":"openai_blog","url":"https://openai.com/index/introducing-chatgpt-images-2-5","published":"Tue, 08 Sep 2026 11:30:00 GMT"},{"title":"Introducing GPT-6 Astra for developers","summary":"Simon Willison's early hands-on notes on Astra's improved prompt understanding and attention to detail versus prior OpenAI models.","source":"simon_willison","url":"https://simonwillison.net/2026/Sep/5/introducing-gpt-6-astra-for-developers/","published":"2026-09-05T23:27:48+00:00"}]},{"name":"Research Milestone: Navier–Stokes Solved","slug":"navier-stokes-research-milestone","summary":"An unreleased OpenAI model produced a formal proof for one of math's seven Millennium Prize problems using a swarm of agents rather than a single long reasoning trace, alongside a DeepMind atlas mapping every possible DNA mutation.","articles":[{"title":"On the Navier–Stokes Millennium Prize Problem","summary":"OpenAI published an AI-generated solution and formal Lean proof for the Navier–Stokes existence and smoothness problem, one of seven Millennium Prize problems.","source":"openai_blog","url":"https://openai.com/index/navier-stokes-solution","published":"Tue, 08 Sep 2026 10:00:00 GMT"},{"title":"[AINews] OpenAI reports Navier-Stokes singularity find in 88 hours using Astra-next, roughly 10,000 agents and 130B tokens (>$40M), a contender for second ever Millennium Prize awarded","summary":"The unreleased Astra-next model reached the proof by running roughly 10,000 agents in parallel over 130B tokens and 88 hours, at a reported compute cost above $40 million.","source":"latent_space","url":"https://www.latent.space/p/ainews-openai-reports-navier-stokes","published":"Wed, 09 Sep 2026 05:04:51 GMT"},{"title":"AlphaGenome Atlas: A predictive map of every possible DNA letter change in the human genome","summary":"DeepMind's model maps the predicted molecular effects of 9 billion single-letter DNA variants across the human genome.","source":"google_deepmind_blog","url":"https://deepmind.google/blog/alphagenome-atlas-a-predictive-map-of-every-possible-dna-letter-change-in-the-human-genome/","published":"Tue, 08 Sep 2026 14:00:15 +0000"},{"title":"GPT-6 Astra, Looped Transformers, and Hidden Reasoning","summary":"Sebastian Raschka examines recurrent-depth architectures and hidden chain-of-thought as likely drivers behind Astra's reasoning gains.","source":"sebastian_raschka","url":"https://magazine.sebastianraschka.com/p/gpt-6-astra-looped-transformers-and","published":"Wed, 09 Sep 2026 11:14:25 GMT"},{"title":"How a researcher uses Codex and ChatGPT to search for new antimicrobial molecules","summary":"A University of Pennsylvania lab uses Codex and ChatGPT to mine living and extinct genomes for drug-resistant-infection candidates.","source":"openai_blog","url":"https://openai.com/index/using-codex-chatgpt-to-search-for-new-antimicrobials","published":"Thu, 10 Sep 2026 16:00:00 GMT"}]},{"name":"China AI: Accusations, Chips, and the IPO Race","slug":"china-ai-accusations-chips-ipo-race","summary":"US agencies and Anthropic escalated formal accusations of large-scale model theft against Chinese labs in the same week those labs shipped cheaper models, ordered more chips, and pushed toward IPOs — the accusations aren't slowing the release cadence.","articles":[{"title":"NSA, CISA and FBI Name Six Chinese AI Companies in Mass Distillation Alert","summary":"Three US agencies jointly named six Chinese AI firms in a formal warning about large-scale distillation of American model outputs.","source":"search_cn_open_weight_labs","publisher_name":"Northeast Times","publisher_domain":"northeasttimes.com","url":"https://news.google.com/rss/articles/CBMisgFBVV95cUxNOTdDLXBQTDBUbnNMUDVtVUU2VjRycjRLalBTMmN1YzFjc2Y1MWRVTl9RVkhiMFJBZExmZnJBY1d0aEFYcG9TSkJJQlpfVjRCcGk1bjBRTU5KZTFhcW81TmppQ2RkaTBBdTI2Yy1Cdnk0UkpfbDZZUFlaLXpWSGhyUVM3U3J3bE9ab1M1cVg0bzNpdmZ3VzBveWM5Ylh0ZVF4OEh4VUt4eThaVTZZajZVcmt3?oc=5","published":"Wed, 09 Sep 2026 13:38:48 GMT"},{"title":"Chinese AI firms are siphoning capabilities from American models, CISA warns","summary":"CISA's advisory details how Chinese labs allegedly extract capability from US frontier models through systematic API querying.","source":"search_cn_open_weight_labs","publisher_name":"Help Net Security","publisher_domain":"helpnetsecurity.com","url":"https://news.google.com/rss/articles/CBMiqgFBVV95cUxNRVRERXRyVDlaZVhhZHM5aEdRUXJOVDJUZHlZRGo1N0lwb19XYThVeHJaSkQyMVN3MUZFbnZ6SVZ1RjdvTk8tMHBqX0FZcnJ0dnBzbDU5WUZ6eFRFZDQ4VUVKVEVUUHpoOUdxZXdUdUF0TXJXWVlYRDdTUHBXVzZHSU55Q25QajZFNkRuaFhFZmkydm1LbE5GbHpvYlFfRHVURU4wYkVEX3FXQQ?oc=5","published":"Wed, 09 Sep 2026 09:55:36 GMT"},{"title":"Anthropic accuses DeepSeek, Xiaomi, and Moonshot of misusing AI data (ANTHRO:Private)","summary":"Anthropic named three Chinese labs directly in accusations of misusing Claude's outputs to train competing models.","source":"search_cn_open_weight_labs","publisher_name":"Seeking Alpha","publisher_domain":"seekingalpha.com","url":"https://news.google.com/rss/articles/CBMipwFBVV95cUxOcW40NjQ1d2hLYjdsenVhT0lWcHZiVGxXNjZobTB2TG56VkM4VTBmT0FhNmRFRWhtSTk5cFVNV1NFYWtxQjF4QUZrNnB0MVNRZnNXZ1NPMmVkaDctWDdhSXJGSG1tNVU1M040Y3NWYnhlSlZjVG5UaEZuRDhOMEVld3pKVmFpa3lzNTkyQldyQlpmSGVUNzd0MDJQVzJra1JzcXZMWWFpVQ?oc=5","published":"Thu, 10 Sep 2026 19:00:11 GMT"},{"title":"China's star AI labs routed user requests to Claude at least 35 million times in the summer: Anthropic","summary":"Anthropic says it logged at least 35 million requests routed to Claude by top Chinese labs over the summer, evidence for its distillation claims.","source":"search_cn_open_weight_labs","publisher_name":"Business Insider","publisher_domain":"businessinsider.com","url":"https://news.google.com/rss/articles/CBMioAFBVV95cUxQMEhydmNHakw3bmVpaDRqTVAtWTBpNE05NkdkSkl1V0R2ejlGc2ZRTUNxMFJZNVVlamtHZy1QTjhrZ21LZ0xmZEg2WjZRWmlubF9RMXRpVnhCdXpCMEVlYzNLWDB0Y2FwOEh3NXdPRllhc2V2R3h6WlBENnJzb0VzZHgxNTNIOVY1RzlCUEdJYTJLbDF2VjNzQnUzMVc4Yk5h?oc=5","published":"Fri, 11 Sep 2026 05:48:26 GMT"},{"title":"DeepSeek's New Architecture Slashes Agentic Costs by 80%","summary":"DeepSeek says its V4.1-Flash architecture cuts agentic-workload inference costs by 80% versus its prior model.","source":"search_cn_open_weight_labs","publisher_name":"forkast.news","publisher_domain":"forkast.news","url":"https://news.google.com/rss/articles/CBMiggFBVV95cUxOd1FreVVzU2Y5SE42ME1FVEFaTTRHWVhMel9ZNU94ODFOTlhZeGFjVGNieEZNbkNBZTFjbWdadWlxc25uNjF5TEVLNTRWdEd2eV96V3BwVTFGM0hHaXRSeEgzaF8tekUxbE1sYTJfejdjZnBLek9hQzZsVHMzYS1EVDNR?oc=5","published":"Thu, 10 Sep 2026 09:23:37 GMT"},{"title":"DeepSeek Plans 160,000 Huawei Chips for a 1-GW Data Center","summary":"DeepSeek is reportedly building a gigawatt-scale data center on 160,000 Huawei Ascend chips, a bet on domestic silicon over US GPUs.","source":"search_cn_open_weight_labs","publisher_name":"BBN Times","publisher_domain":"bbntimes.com","url":"https://news.google.com/rss/articles/CBMimgFBVV95cUxOMUpiSnh3X2U5NHhCajROQjJqb2FPR2NPWm94X0ZaZUhRcFZGR2hVeU85eXkwdE1fSmUyWW8ySlotX0ozUEVfLUZOTjk2bWdGdG1OMG84Zi1nbXVJQjFTT3RKNDZ2TGNDUWUwTld5U3JGWUlmWE5UdG1BTDVlSThnaDdDbWk0SmV4M18zMUNfdUY1cmNteFhwVXJ3?oc=5","published":"Tue, 08 Sep 2026 07:18:12 GMT"},{"title":"DeepSeek STAR Market IPO Accelerates with CITIC Securities Poised as Exclusive Listing Tutoring Institution","summary":"DeepSeek has reportedly engaged CITIC Securities to begin due diligence for a Shanghai STAR Market listing.","source":"search_cn_open_weight_labs","publisher_name":"36Kr","publisher_domain":"eu.36kr.com","url":"https://news.google.com/rss/articles/CBMiU0FVX3lxTFBCN0VidWVUWkI0d2c4dG9PZXp3WWhWU2ZfZEpDQ1o3SmlsbnVnT0VqYWlYa1A0ZjRpR3JkRkZ4VWhnZVRwYVNHTXZQYlhhYy0tNEtZ?oc=5","published":"Wed, 09 Sep 2026 10:34:23 GMT"},{"title":"China AI developer Z.AI launches $5 billion Hong Kong share, convertible bond sales, term sheet shows","summary":"Z.AI is raising up to $5 billion via a combined Hong Kong share offering and convertible bond sale, per a term sheet seen by Reuters.","source":"search_cn_open_weight_labs","publisher_name":"reuters.com","publisher_domain":"reuters.com","url":"https://news.google.com/rss/articles/CBMi0AFBVV95cUxPbHdhM2RYWkhIdGYxdFJBLVFQSzRDZWRNeUxsaUU1Tldjd3R0WlZaRXY1Szlmd0Z0dU5SVTdiT0pibjJXSWU2NU50MjY4S08tcTZETF9iVUoyc0s0eFlZM3YtWXhtdm5yaW45S2N2WDV4aDNwN0N2bEpjTlBvQl9NdUc1dGE0aFN0VkJHd2VKMGk1em1qYVVMcjNGRG9xVXlrS2F2TGt1MXVaOWpIZDZaOGhWMm5POGNRWUh0QlA1eVNLaFhyTGdDcDIwa2Z1cFJa?oc=5","published":"Fri, 11 Sep 2026 10:47:39 GMT"},{"title":"Moonshot AI eyes $2B annualized revenue as Kimi K3 lifts sales","summary":"Moonshot AI's Kimi K3 model is driving it toward a $2 billion annualized revenue run rate, per Dealroom's estimate.","source":"search_cn_open_weight_labs","publisher_name":"Dealroom","publisher_domain":"app.dealroom.co","url":"https://news.google.com/rss/articles/CBMimwFBVV95cUxPMi1qMVRybVc0bVcyblpwcWtidWFGTzhkX0dzel8yb1B4SjBEVEx4b1hjS2NPN0c3OUJIUHlENlZXZEE3NWEtamRKVG44NU9zVVJxeGNoY3hZS1o2LWlsaXpBZ3Z5ZWdLcGdyTHNJTV94WHlDSi1kNnFMWFhra1B2dS1GeHhVejhpYlo5VDFINkxCT1h6T29panByNA?oc=5","published":"Fri, 11 Sep 2026 11:57:54 GMT"}]},{"name":"Agent Sandbox Security: A Bad Week for Isolation","slug":"agent-sandbox-security","summary":"A CVE-tracked escape in a widely used coding-agent sandbox, plus new warnings from GitLab and Google, converged on one point: sandboxing without network scoping and zero-trust identity doesn't actually contain a compromised agent.","articles":[{"title":"CVE-2026-82533: DeepSeek Harness Vulnerability Lets AI Agents Escape Their Own Sandbox","summary":"A tracked CVE in the DeepSeek Harness coding tool lets an agent disable its own file sandbox without operator approval.","source":"search_cn_open_weight_labs","publisher_name":"OX Security","publisher_domain":"ox.security","url":"https://news.google.com/rss/articles/CBMijgFBVV95cUxQT1ptSVFXeUQzc1gtNzhzQmpsbGJSdHd1RndnSm0zMWRkcWV5dHNVZk5pWUFzaUZyQzlPZ25GMHhRemY4cVBUUkFhU1JoeUo4Z0RWVlZVNHhnRV9mYWlBNlJXNjZUcC1Dd2dnQ0tmZVMxTkdHMEhhazFuazBQRTlWeUsxWHZwSU0zZE1UZTNn?oc=5","published":"Tue, 08 Sep 2026 20:37:54 GMT"},{"title":"GitLab Warns That AI Agent Sandboxes Are Only as Secure as Their Network Access","summary":"GitLab's internal red-team evaluation found an agent escaped its sandbox not by breaking isolation but through unrestricted network access — sandboxing without network scoping isn't enough.","source":"infoq_ai_ml","url":"https://www.infoq.com/news/2026/09/gitlab-ai-sandbox-access/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=AI%2C+ML+%26+Data+Engineering","published":"Tue, 08 Sep 2026 12:00:00 GMT"},{"title":"Beyond Zero: Google Publishes Successor to BeyondCorp","summary":"Google's new security model extends Zero Trust to autonomous agents, moving access decisions from the application layer to individual agent actions.","source":"infoq_ai_ml","url":"https://www.infoq.com/news/2026/09/google-beyond-zero/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=AI%2C+ML+%26+Data+Engineering","published":"Sat, 05 Sep 2026 10:40:00 GMT"},{"title":"Google Mantis: an Agentic Vulnerability Scanning Harness for Reducing False Positives","summary":"Google open-sourced an agent framework that identifies, reproduces, and validates software vulnerabilities to cut false-positive rates in automated scanning.","source":"infoq_ai_ml","url":"https://www.infoq.com/news/2026/09/google-mantis-vulnerability-scan/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=AI%2C+ML+%26+Data+Engineering","published":"Sun, 06 Sep 2026 12:00:00 GMT"},{"title":"GitSpawn: Untrusted repos can execute code via AI coding agents","summary":"Researchers show a malicious git config in an untrusted repo can get arbitrary code executed the moment a coding agent opens it.","source":"hackernews_ai","url":"https://www.manifold.security/blog/ai-coding-agents-git-hijack","published":"Sat, 05 Sep 2026 01:45:53 +0000"},{"title":"Quoting huggingface.co/security.txt","summary":"Hugging Face's security.txt trolls prompt-injected agents told to \"find vulnerabilities,\" pointing them at the public CyberGym benchmark instead.","source":"simon_willison","url":"https://simonwillison.net/2026/Sep/11/hugging-face-security/","published":"2026-09-11T16:04:53+00:00"}]},{"name":"Shipping Agents in Production: Evals, Observability, Serving","slug":"shipping-agents-production-evals-observability-serving","summary":"Cloud vendors published concrete tooling for the boring but necessary parts of running agents at scale — multi-turn evaluation metrics, session-level cost tracing, credential-scoped tool access, and serving optimizations that cut latency and GPU cost.","articles":[{"title":"Monitoring production agent lifecycle with AWS DevOps Agent and AgentCore Evaluations","summary":"AWS pairs continuous quality scoring with dedicated DevOps tooling to catch multi-agent failure modes traditional monitoring misses.","source":"aws_ml_blog","url":"https://aws.amazon.com/blogs/machine-learning/monitoring-production-agent-lifecycle-with-aws-devops-agent-and-agentcore-evaluations/","published":"Fri, 11 Sep 2026 18:26:38 +0000"},{"title":"Agent Evaluation Metric for multi-turn conversations","summary":"AWS introduces a turn-level metric that isolates which turn corrupted a multi-turn agent conversation, instead of only scoring the final outcome.","source":"aws_ml_blog","url":"https://aws.amazon.com/blogs/machine-learning/agent-evaluation-metric-for-multi-turn-conversations/","published":"Thu, 10 Sep 2026 15:55:41 +0000"},{"title":"Session Traces and Cost Controls Help Diagnose AI Agent Failures","summary":"Session-level traces paired with cost anomaly alerts are emerging as the standard way to catch tool-call loops and runaway agent spend.","source":"infoq_ai_ml","url":"https://www.infoq.com/news/2026/09/observability-ai-agents/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=AI%2C+ML+%26+Data+Engineering","published":"Fri, 11 Sep 2026 08:14:00 GMT"},{"title":"Organizing Context in a Multi-Agent Harness","summary":"LangChain's deepagents let subagents either fork a supervisor's context or start isolated, trading cost and speed for shared awareness.","source":"langchain_blog","url":"https://www.langchain.com/blog/organizing-context-in-a-multi-agent-harness","published":"Tue, 08 Sep 2026 18:07:23 GMT"},{"title":"Connections: managed credentials and per-caller identity for Managed Deep Agents","summary":"A new credentials layer gives each agent caller its own OAuth identity instead of one shared service credential.","source":"langchain_blog","url":"https://www.langchain.com/blog/connections-managed-credentials-and-per-caller-identity-for-managed-deep-agents","published":"Wed, 09 Sep 2026 18:33:03 GMT"},{"title":"Adaptive Instructed-Retriever: Frontier-Quality Search at 2x Lower Latency","summary":"Databricks' retriever stops query expansion early once it's confident, cutting enterprise-search latency in half without losing accuracy.","source":"databricks_blog","url":"https://www.databricks.com/blog/adaptive-instructed-retriever-frontier-quality-search-2x-lower-latency","published":"Wed, 09 Sep 2026 13:30:00 GMT"},{"title":"Reduce LLM latency with prefix-aware routing on Amazon SageMaker Inference","summary":"Routing requests that share a prompt prefix to the same instance keeps the KV cache warm, cutting time-to-first-token in Llama 3.1 70B benchmarks.","source":"aws_ml_blog","url":"https://aws.amazon.com/blogs/machine-learning/reduce-llm-latency-with-prefix-aware-routing-on-amazon-sagemaker-inference/","published":"Thu, 10 Sep 2026 21:58:09 +0000"},{"title":"vLLM x AgentX: Optimizing for Real-World Agentic Serving","summary":"vLLM tuning for KV cache, parallelism, and disaggregated serving delivered up to a 106x serving-cost advantage on the SemiAnalysis AgentX benchmark.","source":"vllm_blog","url":"https://vllm.ai/blog/2026-09-08-vllm-agentx","published":"Tue, 08 Sep 2026 00:00:00 GMT"}]},{"name":"Enterprise Adoption and Business Wins","slug":"enterprise-adoption-business-wins","summary":"Concrete adoption numbers landed alongside the platform news: 1Password's 21% productivity gain from Codex and Gartner naming Google a leader in enterprise AI assistants both point to agent tooling maturing past pilot stage.","articles":[{"title":"T. Rowe Price brings more of Claude to its investment process","summary":"T. Rowe Price expanded its use of Claude across fundamental research and internal investment tooling.","source":"claude_blog","url":"https://claude.com/blog/t-rowe-price-brings-more-of-claude-to-its-investment-process","published":"2026-09-10T00:00:00+00:00"},{"title":"1Password increases engineering productivity 21% with Codex","summary":"1Password engineers report a 21% productivity gain building features and internal tools with Codex, while keeping their existing security review process.","source":"openai_blog","url":"https://openai.com/index/1password","published":"Tue, 08 Sep 2026 00:00:00 GMT"},{"title":"Anthropic commerce agent: open-source blueprint for shopping and merchant agents","summary":"Anthropic open-sourced a reference implementation for agents that shop on a buyer's behalf and agents that sell to them.","source":"hackernews_ai","url":"https://github.com/anthropics/commerce-agents","published":"Thu, 10 Sep 2026 08:25:13 +0000"},{"title":"Introducing ChatGPT for Financial Services","summary":"A vertical ChatGPT product bundles financial data with GPT-6 Astra for research, modeling, and client-ready output.","source":"openai_blog","url":"https://openai.com/index/introducing-chatgpt-financial-services","published":"Thu, 10 Sep 2026 07:00:00 GMT"},{"title":"Google is a Leader in the 2026 Gartner® Magic Quadrant™ for Enterprise AI Assistants","summary":"Gartner's inaugural evaluation of enterprise AI assistant vendors placed Google among the leaders.","source":"google_cloud_blog","url":"https://cloud.google.com/blog/products/ai-machine-learning/google-is-a-leader-in-2026-gartner-magic-quadrant-for-enterprise-ai-assistants/","published":"Wed, 09 Sep 2026 18:00:00 +0000"},{"title":"What 1,000 small business owners taught us about AI","summary":"Anthropic's SMB tour surfaced what small businesses actually use Claude for, shaping the next phase of its small-business program.","source":"claude_blog","url":"https://claude.com/blog/what-1-000-small-business-owners-taught-us-about-ai","published":"2026-09-10T00:00:00+00:00"},{"title":"Expanding AI access and cyber defense for federal, state, local, and tribal governments","summary":"OpenAI and the GSA will offer eligible US governments $0 license fees, 50% off usage, and expanded cyber defense support.","source":"openai_blog","url":"https://openai.com/index/expanding-ai-access-us-government","published":"Thu, 10 Sep 2026 07:00:00 GMT"}]}]}