Trusted AI Blog

This week in AI security: chatbots like ChatGPT have become essential online tools, and their security loopholes are an emerging concern.

The recent DefCon hacker conference in Las Vegas served as a stark reminder of the pressing concerns around AI safety and security.

With the White House's endorsement, the Generative Red Team Challenge aims to rigorously assess the reliability of AI.

This week in AI security: the generative AI contest between companies and hackers, securing LLM systems against prompt injection, and Google's AI red team.

• LLM Security

A digest for anyone learning prompt engineering, or building with LLMs and trying to understand the security issues that come with them.

• LLM Security

Over the past eight months, OpenAI's ChatGPT has dazzled millions with its ability to produce lifelike text, from stories to code.

Ethical AI Database (EAIDB) has included Adversa AI Security for the AI category.

• Adversarial ML

Four papers on AI offense and defense: vulnerabilities in NLI models and generative AI, adversarial attacks on 3D point clouds, and misuse of multimodal LLMs.

This week in AI security: the Biden-Harris administration's commitment to responsible and secure AI, and the global initiatives moving the same direction.

• LLM Security

This week in AI security: an AI detector mislabeling non-native English writing, the dark side of AI, and the real challenges of enterprise LLM adoption.

This week in AI security: the CISO role transforming as organizations take on greater technological needs and risks, and where the secure AI market goes.

• LLM Security

We launched a new newsletter dedicated to LLM Security.