RELIC LAW PLLC

AI systems are rapidly being deployed in a variety of contexts. However, AI systems give rise to a multitude of risks and compliance challenges that are often underappreciated by AI system owners and operators. From data processing performed by AI systems regulated as "automated decision making" to unlawful discrimination to emerging AI laws and regulations, organizations must fully evaluate these risks before putting new AI systems into production. 

​

Whether you are developing a new AI system or leveraging a "pre-trained" model, the AI system must be thoroughly assessed prior to entering production to avoid regulatory failures and risks associated with undesirable outputs. It is widely understood that most AI systems suffer from biases linked to their training and/or training data. At scale, low confidence intervals and high error rates in outputs that exceed acceptable tolerances can also lead to serious risk and exposure to liability.  

 

As with any other system, AI systems must be evaluated so that unacceptable risks can be eliminated and any remaining risks can be properly treated by the AI system owner or operator. AI systems that run afoul of legal or regulatory restrictions must be retooled or re-engineered to operate within regulatory limits. Because the requirements of different jurisdictions may be asymmetrical, a global compliance strategy may be necessary. 

 

Relic Law employs a multi-step approach to evaluate AI system risk and compliance. Our AI system playbook incorporates all of the latest legal and regulatory guidance as well as industry standards and best practices. This includes feedback and recommendations on AI system design, such as weights, data sourcing, and labeling. In addition, our multidisciplinary team will assess critical privacy, cybersecurity, and intellectual property risks that may arise from the AI system or its use. 

 

Our team is ready to assist you in managing your AI system risk and compliance.

 

Let's discuss your AI strategy. (248) 579-9537