AI Daily — May 19, 2026
2026-05-19
anthropic.com
Anthropic acquires Stainless
Anthropic has acquired Stainless, the SDK generation startup that automatically builds and maintains client libraries from OpenAPI specs. The acquisition likely strengthens Anthropic's developer tooling and API ecosystem, making it easier for developers to integrate Claude across languages and platforms. This is a notable infrastructure play, signaling that Anthropic is investing in the developer experience layer around its API.
openai.com
OpenAI and Dell partner to bring Codex to hybrid and on-premise enterprise environments
OpenAI and Dell are partnering to deploy Codex — OpenAI's AI coding agent — in hybrid and on-premise enterprise environments, targeting organizations with strict data residency and security requirements. The partnership allows enterprises to run Codex agents against internal codebases and workflows without routing sensitive code through OpenAI's cloud infrastructure. This extends the competitive reach of AI coding agents into regulated industries that have previously been locked out of cloud-only deployments.
technologyreview.com
Here's why Elon Musk lost his suit against OpenAI
A jury delivered a unanimous advisory verdict against Elon Musk in Musk v. Altman, ruling his claims are barred by the applicable statutes of limitations — which Judge Yvonne Gonzalez Rogers immediately accepted. The decision is a significant legal setback for Musk's challenge to OpenAI's for-profit conversion. Musk has announced plans to appeal.
huggingface.co
The Open Agent Leaderboard
IBM Research and Hugging Face have launched the Open Agent Leaderboard, a public benchmark for evaluating LLM-based agents on real-world agentic tasks using open evaluation infrastructure. The leaderboard focuses on reproducibility and transparency, targeting a gap left by proprietary agent evaluations. This provides a standardized reference point for comparing agent systems across open and closed models.
arxiv.org
The Scaling Laws of Skills in LLM Agent Systems
Researchers identify two coupled scaling laws governing LLM agent systems as skill libraries grow: routing accuracy decays logarithmically with library size (R²>0.97 across 15 frontier LLMs and 1,141 skills), and execution outcomes compound multiplicatively before state realization. A single parameter — the routing logarithmic decay slope — couples both laws, enabling predictions about system degradation at scale. The findings have practical implications for designing large-scale skill-based agent architectures and anticipating where routing failures emerge.