📖 2 min read
Friday, May 1, 2026 — your fast-read digest of what moved in AI overnight.
OpenAI begins GPT-5.5 rollout to ChatGPT and Codex
OpenAI started pushing GPT-5.5 to Plus, Pro, Business, and Enterprise tiers in ChatGPT and Codex, with GPT-5.5 Pro rolling out to Pro, Business, and Enterprise users. API access is being staged separately while OpenAI works through deployment safeguards with partners.
📧 Want more like this? Get our free The Ultimate AI Tool Database: 200+ Tools Rated & Ranked — Downloaded 5,000+ times
This is the first major capability bump since the GPT-5 line launched, and Codex users get it day-one — a clear signal that coding agents remain the company’s flagship use case. If you’re benchmarking models for your stack, our team keeps a running scorecard at AiToolCrush.
Anthropic launches Claude Security in public beta
Anthropic moved Claude Security out of limited research preview and into public beta for Claude Enterprise customers. The product scans codebases for vulnerabilities and ships with a confidence rating system designed to kill the false-positive problem that plagues most security tooling.
According to Anthropic, internal pilots compressed days of back-and-forth between security and engineering teams into a single sitting. It’s the same Claude Code engine that’s been battle-tested over the last few months, now pointed squarely at the AppSec market.
📧 Want more like this? Get our free The Ultimate AI Tool Database: 200+ Tools Rated & Ranked — Downloaded 5,000+ times
Google rolls Gemini into millions of GM vehicles
Google announced that Gemini is replacing Google Assistant in cars with Google built-in, starting with a deployment to roughly 4 million GM vehicles — no dealer visit required. Drivers get hands-free messaging, conversational navigation, and contextual answers from the dashboard.
It’s one of the largest single-day AI deployments to consumer hardware so far this year, and it lands the same week Google introduced unique agent identities for Gemini Enterprise, signaling a serious push on agentic AI across both consumer and B2B surfaces.
Nous Research drops Hermes Agent framework
Nous Research published hermes-agent, an open-source agent runtime that plugs into virtually any model backend — Nous Portal, OpenRouter, NVIDIA NIM (Nemotron), Xiaomi MiMo, z.ai/GLM, Kimi/Moonshot, MiniMax, Hugging Face, OpenAI, or self-hosted endpoints. The pitch: an agent that grows with you, model-agnostic by design.
📧 Want more like this? Get our free The Ultimate AI Tool Database: 200+ Tools Rated & Ranked — Downloaded 5,000+ times
For the LocalLLaMA crowd this is a meaningful unlock — autonomous workflows without lock-in to any single provider. Expect quick integrations with the open-weight tier (GLM 5.1, Qwen, DeepSeek) over the weekend.
Gen launches “VPN for Agents” — a first for consumer AI
Gen (the Norton/Avast parent) unveiled VPN for Agents, billed as the first consumer AI-native VPN purpose-built for autonomous AI agents, available through the new Gen Agent Trust Hub. The product addresses a problem that’s been quietly building: agents browsing the web on your behalf inherit your IP, your fingerprint, and your risk profile.
As more people run always-on agents for research, monitoring, and trading workflows — including prediction-market sweeps like the ones we cover at BetOnAI — separating the agent’s network identity from the human’s is going to become table stakes.
That’s the wrap for May 1. Back tomorrow with the next batch.