Here are today's news for 23rd of October about IBM, Anthropic, OpenAI and Google. Let's dive into the details. On Oct. 20, 2026, contributors announced llm, d 1.0 at the PyTorch Conference North America, hardening the open, source, Kubernetes, native framework for production, scale AI inference. Founded in 2025 by IBM Research, Red Hat, Google Cloud, CoreWeave and NVIDIA, the project now has more than 500 contributors across 25, plus organizations and joined the Cloud Native Computing Foundation as a sandbox project in March 2026. Version 1.0 adds priority, and fairness, aware multi, tenant traffic control, router hot standby and rolling upgrades, coordinated management of prefill and decode replicas, plus session, aware routing for agent conversations. It works with vLLM and SGLang, across GPUs and Google TPUs, and supports prefix, cache routing, KV, cache offloading and distributed model serving. IBM Research reports a roughly 98% prefix, cache hit rate and about $0.11 per million tokens in one self, hosted deployment. In another demonstration, llm, d served an approximately 753, billion, parameter GLM 5.2 model at more than 6. On October 22, 2026, Anthropic reported that Claude, working with software, verification firm Theorem, proved crash, safety properties for 239 test entry points across 123 open, source libraries from Google’s OSS, Fuzz. Over several weeks, Claude translated existing fuzz tests into formal specifications, patched bugs it uncovered, and generated machine, checkable proofs covering every input under 64 MB. The proofs apply to compiled RISC, V binaries and establish normal exit without memory, safety errors or failed assertions, subject to assumptions including single, threaded execution, static linking, a fixed memory budget, and modeled Linux behavior. The work produced about 180 million lines of Rocq proof code; a typical proof was roughly 600,000 lines. It exposed hundreds of bugs, including a decompressor counter underflow enabling an out, of, bounds write on a rare, roughly 20 MB input. On October 21, 2026, OpenAI and Google Cloud announced that OpenAI’s GPT, 6 Astra, GPT, 6.1 Sol and GPT, 6 Luna, plus Codex agentic workflows, are coming to Google Cloud’s Model Garden. The three models are rolling out in private preview to selected enterprise customers; general availability is planned in the coming weeks. Customers will buy through Google Cloud Marketplace, with eligible existing GCP commitments usable for purchases and usage consolidated on Google Cloud invoices. Google says the models span demanding coding, research and professional work, lower, cost agentic tasks, and fast everyday workloads; no prices or benchmark scores were announced. For regulated businesses, the appeal is access to OpenAI models alongside Google’s access controls, governance and infrastructure. The deal also reaches Google’s Antigravity development environment and Gemini Agent, while giving OpenAI distribution across all three major US cloud providers. On October 22, 2026, Google Cloud announced Agent Compute Engine at AGNTCon + MCPCon, with customer preview planned for the coming months. The service gives each AI agent a hardware, isolated microVM, designed to start quickly while providing its own guest kernel and root access. Google says the system already powers Gemini agent workloads and internal evaluations; it has not announced pricing or CPU and memory limits. The security design assumes an agent’s sandbox could be compromised: credentials stay outside the microVM, while outbound requests pass through Agent Gateway for identity checks and policy enforcement. Built on Cloud Hypervisor and Google’s Titanium infrastructure, the service can checkpoint idle sessions, freeing active compute while preserving files and memory for resumption. Google says this lets a physical node support “orders of magnitude” more living agent microVMs than conventional VMs, but published no benchmark for that claim. The launch extends Google’s effort to make agents usable at enterprise scale. Federal disclosures filed in October show Nvidia, Anthropic and OpenAI each set a quarterly lobbying record from July through September 2026. Nvidia spent about $2.2 million, Anthropic $2.1 million and OpenAI $1.5 million, roughly $5.8 million combined. The Verge reported that Nvidia’s figure was its first above $2 million in a quarter; the total for all three was up 22% from Q2 and nearly 60% year over year, according to POGO’s review. Their filings cover efforts to influence policy, not spending on a single bill. Nvidia listed issues including export controls, competition policy, responsible AI design, quantum computing and 6G. The surge signals that AI policy has become a costly competitive arena: chip access and rules for deploying models can shape companies’ markets, while lawmakers and agencies face pressure from firms with sharply different interests. The spending follows rapid growth in Washington: POGO says the three have spent about $37 million on federal lobbying since 2024. Nvidia’s 2026 total had already reached $6. That's all for today's AI news briefing. Have a great day!