AIが期間内の動向を整理
OpenAI adds Sites to Codex, Anthropic expands vulnerability discovery, and AI development and operations move forward
The AI news of June 3 was marked by moves directly relevant to development teams and security operations. OpenAI added plugins and Sites to Codex, Anthropic expanded Project Glasswing to accelerate vulnerability discovery, and Google clarified the positioning of the Gemini 3.5 family with a stronger focus on task-based usage.
SOURCE CHECK
Primary Sources 13
Primary Sources
SOURCE CHECK 1
openai.com
SOURCE CHECK 2
x.com
SOURCE CHECK 3
x.com
SOURCE CHECK 4
x.com
SOURCE CHECK 5
openai.com
SOURCE CHECK 6
anthropic.com
SOURCE CHECK 7
anthropic.com
SOURCE CHECK 8
anthropic.com
SOURCE CHECK 9
x.com
SOURCE CHECK 10
anthropic.com
SOURCE CHECK 11
deepmind.google
SOURCE CHECK 12
deepmind.google
SOURCE CHECK 13
anthropic.com
Key Points
- 1OpenAI expanded Codex with specialist tools and Sites to shorten workflows
- 2Anthropic broadened Project Glasswing and reported more than 10,000 high- or critical-severity vulnerabilities found
- 3Google highlighted the Gemini 3.5 family and organized its agent and coding use cases
- 4For enterprises, support for moving from evaluation to production continues
OpenAI turns Codex into a work tool
OpenAI expanded Codex plugins and added specialist tools for sales, data analytics, creative work, and more. The Sites feature can turn ideas directly into websites or apps. This is useful not only for development teams, but also for planning and sales teams that want to move quickly from idea to prototype.
Anthropic expands vulnerability discovery efforts
Anthropic expanded Project Glasswing and broadened the activity of early partners using Claude Mythos Preview. More than 10,000 high- or critical-severity vulnerabilities have already been found, making this especially relevant to practical security improvement work. Companies handling important systems will find it easier to judge the value of the initiative.
Google organizes how to use Gemini 3.5
Google DeepMind put the Gemini 3.5 family front and center and clarified its positioning for agents and coding. It also points builders to Google AI Studio, the Gemini API, and Google Antigravity, creating a clearer path from testing to implementation for developers.
What enterprise users should watch
What these three items have in common is that AI is becoming more structured for real business use. The focus is shifting toward where it helps: creating, securing, and deploying. For information gathering, the key is not just what is new, but how it connects to your own workflows.