Anthropic released Claude Opus 4.6, its new flagship model, which featured a one million token context window and provided more reliable retrieval of relevant information in large documents than previous Opus models.
2.
OpenAI launched Frontier, a platform that gave AI agents employee-like identities, shared context, and enterprise permissions, and that was rolled out first with selected enterprise customers.
3.
OpenAI's new coding model GPT-5.3-Codex set new highs on agentic coding benchmarks and, according to the company, helped build itself during training and deployment.
4.
Cerebras Systems closed a financing round of over $1 billion, valuing the company at about $23 billion, and reported a recent $10 billion deal with OpenAI.
5.
OpenClaw's OpenDoor vulnerability was shown by security researchers to allow attackers to take complete control through manipulated documents, enabling installation of a permanent backdoor and compromising users' computers.