# Morning AI TL;DR — 2026-05-24

- **DeepSeek makes V4‑Pro’s 75% price cut permanent** — Reuters/Bloomberg report the flagship model will stay at roughly one-quarter of its prior developer price, intensifying API price pressure for agent and app builders. [Reuters via U.S. News](https://money.usnews.com/investing/news/articles/2026-05-23/chinas-deepseek-to-make-permanent-75-price-cut-on-flagship-v4-pro-ai-model) / [Bloomberg](https://www.bloomberg.com/news/articles/2026-05-23/deepseek-to-make-permanent-75-discount-on-flagship-ai-model)
- **Anthropic launches Claude Opus 4.5** — Anthropic says the new model is aimed at coding, agents, computer use, research, slides, and spreadsheets, with GitHub Copilot testing pointing to strong coding efficiency. [Anthropic](https://www.anthropic.com/news/claude-opus-4-5?ref=lmarena.ai)
- **OpenAI’s GPT‑5.4 push centers on computer-use agents** — OpenAI positions GPT‑5.4 in ChatGPT, API, and Codex as a work model for long workflows across apps, which matters for builders wiring agents into real desktop/browser tasks. [OpenAI](https://openai.com/index/introducing-gpt-5-4/)
- **Google I/O’s agent stack is still the week’s biggest dev platform story** — Google’s official I/O roundup highlights Gemini Omni, Antigravity, and other agentic interfaces, while coverage says Gemini 3.5 Flash becomes default in Gemini/Search AI Mode. [Google](https://blog.google/innovation-and-ai/technology/ai/google-io-2026-all-our-announcements/) / [The Verge](https://www.theverge.com/tech/933415/google-io-2026-biggest-announcements-ai-gemini)
- **Enterprise AI coding agents are becoming a formal market category** — Gartner says coding agents are shifting software teams from AI-assisted development toward agentic SDLC workflows across planning, coding, and review. [Gartner](https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-says-the-market-for-enterprise-ai-coding-agents-is-entering-a-new-phase-of-expansion-and-competitive-realignment)
- **Bank AI adoption is moving from pilots to workforce impact** — Reuters reports HSBC told staff not to “fight AI” as banks begin job cuts, while Standard Chartered’s CEO apologized after comments about AI replacing “lower-value” work. [Reuters](https://www.reuters.com/legal/legalindustry/dont-fight-ai-hsbc-ceo-tells-staff-banks-begin-job-cuts-2026-05-20/) / [Investing.com/Reuters](https://www.investing.com/news/stock-market-news/stanchart-ceo-apologises-for-upset-caused-by-ai-comments-4705977)

**Worth watching:** Cheaper frontier-ish APIs plus stronger computer-use/coding agents should make governed, Salesforce-integrated agent workflows more economically viable—but review, permissions, and audit trails become the differentiator.

**Artifact:** `/home/tjonestj3/.hermes/hermes-agent/outputs/ai-news/2026-05-24-ai-news-tldr.md`