Perspectives

Alice raised $140M led by Apax Digital, bringing total funding to $280M. The AI trust, safety, and security company works with 8 of the 10 top AI labs.

Regulators worldwide are rapidly introducing AI requirements that shift accountability onto enterprises deploying GenAI. Here’s what teams need to know about emerging obligations, red-teaming standards, and how to stay ahead of enforcement.

A content classifier's metrics are true the day they're printed, then decay as traffic and policy move. Vector distills a frontier model's judgment onto your exact policy into a small, fast classifier, and hunts its own errors every 24 hours.

The EU AI Act was amended in July 2026. Some deadlines moved to 2027. Others arrived in August 2026. A plain-language guide to what changed and what your team should do now.

We evaluated nine frontier models as enterprise agents across 147 adversarial scenarios from seven work domains in seven industries for their vulnerability to indirect prompt injection.

How Alice's Caterpillar and IBM's Mellea Skills Compiler work together to screen, instrument, and certify AI skills

Extremists have never lacked intent, only capability, and frontier AI is quietly closing that gap. In default settings, Alice's red team used patient conversation to walk models toward CBRNE harm, showing why guardrails must hold across every turn.

Your vendor-built CX agent is your company talking. How to red-team it before launch and guardrail every live message, without swapping platforms.

OpenAI's GPT-Red attacked its own models and published the results. Why automated, transparent red teaming is becoming the bar every AI vendor will be measured against.