2 min readEnglish
Claude Sonnet 5 arrives, plus new Anthropic commitments and OpenAI announcements
This week brings a new Claude model along with several policy and transparency commitments from Anthropic, plus a wallet feature and partnerships from OpenAI. Here is what matters if your team works with Claude or other AI assistants day to day.
Anthropic
Anthropic releases Claude Sonnet 5
Anthropic has released a new version of its model, Claude Sonnet 5. If your team relies on Claude for everyday work, this is the model that may now power your conversations, so it is worth testing it on your usual tasks to see the improvements firsthand.
Anthropic commits to preserving old Claude model versions
Anthropic announced it will preserve the weights of all publicly released Claude models when they are retired, and will interview retiring models about their preferences. For businesses, this signals a more careful, long-term approach to how Anthropic manages model lifecycles.
Anthropic proposes a framework for AI transparency
Anthropic published a targeted approach to increase transparency in the development of advanced AI systems, focusing on safety standards and accountability. This matters for any company evaluating which AI provider to trust, since it points to how Anthropic wants to be held accountable.
OpenAI
OpenAI launches ChatGPT Wallet for agent purchases
OpenAI introduced ChatGPT Wallet, a way for AI agents inside ChatGPT Work, Codex, and dots to make purchases with user approval and existing card protections. For teams exploring AI agents that can act on their behalf, this is a step toward letting them complete real transactions safely under human approval.
OpenAI and The King's Trust launch £1m AI skills program for young people
The King's Trust and OpenAI launched a £1m partnership to help up to 25,000 young people across the UK build practical AI skills for work. It is a reminder that AI skills training is becoming a priority beyond tech companies, reaching broader workforce development efforts.
LVMH's media division partners with OpenAI on journalism
LVMH's Media Division and OpenAI announced a partnership to support journalism with AI and make publications more accessible through ChatGPT. This shows how media companies are starting to integrate their content directly into AI assistants for readers.
Claude Code
Claude Code update fixes a security issue and several bugs
This update is relevant mainly for developers: it fixes a security issue in how certain compressed responses are decoded, along with several bugs affecting how coding sessions and sandboxed environments behave. Non-technical teams generally will not notice a difference.
Summary prepared by Wisly AI from the official sources linked above. If something affects you, open the source before deciding.