OpenAI introduced workspace agents in ChatGPT on April 22, 2026, adding a research preview that lets organizations build shared agents for multi-step work such as reporting, lead outreach and software review.
OpenAI on April 14 unveiled GPT-5.4-Cyber, a variant of its flagship model tuned for defensive cybersecurity work, marking a sharper push into a market where buyers want AI that can support incident response, vulnerability review and day-to-day security operations.
Anthropic previewed a new high-end AI model on April 7, 2026 as part of a security-focused initiative, underscoring the company’s effort to position frontier AI as a tool for enterprise risk management and safer deployment.
