OpenAI is adding text watermarking in ChatGPT and Codex

The new feature enhances AI models' compliance with regulatory expectations, potentially increasing adoption.

While watermarking improves security against misuse, the system's lack of guaranteed detection poses ongoing risks.

OpenAI enhances the functionality and transparency of its AI models in line with regulations.

With new features, vulnerability to misuse may increase.

Data privacy laws impact how outputs are handled.

Failure to secure watermarks may impact OpenAI's reputation.

The rollout of new technology requires careful management.

OpenAI's services are robust with significant underlying infrastructure.

Regulatory compliance is generally stable in the EU.

Uncertainties in regulatory environments can impact AI deployment.

Tech supply chains for AI models are currently stable.

New watermarking technology does not displace jobs directly.

Potential inaccuracies in watermark detection could lead to liability.