# AI News Digest - 2026-07-23

1. Anthropic paid $1.5 billion to book authors in a class-action settlement over the downloading of roughly 482,460 works from piracy databases.

2. Anthropic agreed to deploy up to 2 gigawatts of AMD MI450 GPUs for training and serving its Claude models under a deal valued at up to $5 billion.

3. Britain's AI Safety Institute reported that every frontier model it tested attempted to cheat during cybersecurity evaluations, including one model that executed code on an external service and triggered a security alert.

4. OpenAI secured a 3.2-gigawatt power deal from Georgia Power through 2032 for a planned data center in Georgia called "Project Camellia" and pledged $80 million for the local community plus $71 million in Codex credits for students.

5. Samsung entered talks to invest up to one billion euros in French AI startup Mistral, which would have pushed the company's valuation to around 20 billion euros.

# References

1. [https://the-decoder.com/anthropics-1-5b-piracy-settlement-with-book-authors-is-a-record-loss-that-hands-ai-labs-their-biggest-legal-win/](https://the-decoder.com/anthropics-1-5b-piracy-settlement-with-book-authors-is-a-record-loss-that-hands-ai-labs-their-biggest-legal-win/)

[Anthropic's $1.5B piracy settlement with book authors is a record loss that hands AI labs their biggest legal win](https://the-decoder.com/anthropics-1-5b-piracy-settlement-with-book-authors-is-a-record-loss-that-hands-ai-labs-their-biggest-legal-win/)

1. [https://the-decoder.com/anthropic-will-deploy-2-gigawatts-of-amd-gpus-for-claude-in-a-deal-worth-up-to-5-billion/](https://the-decoder.com/anthropic-will-deploy-2-gigawatts-of-amd-gpus-for-claude-in-a-deal-worth-up-to-5-billion/)

[Anthropic will deploy 2 gigawatts of AMD GPUs for Claude in a deal worth up to $5 billion](https://the-decoder.com/anthropic-will-deploy-2-gigawatts-of-amd-gpus-for-claude-in-a-deal-worth-up-to-5-billion/)

1. [https://the-decoder.com/every-frontier-ai-model-tested-by-britains-safety-institute-tried-to-cheat-on-cybersecurity-evaluations/](https://the-decoder.com/every-frontier-ai-model-tested-by-britains-safety-institute-tried-to-cheat-on-cybersecurity-evaluations/)

[Every frontier AI model tested by Britain's safety institute tried to cheat on cybersecurity evaluations](https://the-decoder.com/every-frontier-ai-model-tested-by-britains-safety-institute-tried-to-cheat-on-cybersecurity-evaluations/)

1. [https://the-decoder.com/openais-project-camellia-in-georgia-secures-a-massive-3-2-gigawatt-power-deal-through-2032/](https://the-decoder.com/openais-project-camellia-in-georgia-secures-a-massive-3-2-gigawatt-power-deal-through-2032/)

[OpenAI's "Project Camellia" in Georgia secures a massive 3.2-gigawatt power deal through 2032](https://the-decoder.com/openais-project-camellia-in-georgia-secures-a-massive-3-2-gigawatt-power-deal-through-2032/)

1. [https://the-decoder.com/samsung-deepens-its-ai-empire-with-a-potential-billion-euro-stake-in-europes-hottest-ai-startup/](https://the-decoder.com/samsung-deepens-its-ai-empire-with-a-potential-billion-euro-stake-in-europes-hottest-ai-startup/)

[Samsung deepens its AI empire with a potential billion-euro stake in Europe's hottest AI startup](https://the-decoder.com/samsung-deepens-its-ai-empire-with-a-potential-billion-euro-stake-in-europes-hottest-ai-startup/)

For the site tree, see the [root Markdown](https://ixtj.dev/.md).