Sen. Mark Warner, D-Va., has urged CEOs of multiple artificial intelligence companies to prioritize security and counter bias as they design and deploy artificial intelligence systems.
In letters sent to technology leaders, Warner requested that the recipients respond to a list of 16 questions with regard to how they address security risks associated with AI development, the senator’s office said Wednesday.
Security risks identified by Warner, who serves as chairman of the Senate Select Committee on Intelligence, include data poisoning attacks, data supply chain security concerns, algorithmic bias, trustworthiness and malicious use of AI systems.
Questions listed in the letter ask about the technology companies’ security approaches, implementation of restrictions on third-party access to AI models, measures to ensure the security and accuracy of data inputs and outputs and efforts to monitor and audit systems to detect and mitigate security breaches.
The letters were addressed to chief executives of Anthropic, Apple, Google, Meta, Microsoft, Midjourney, OpenAI, Percipient.ai, Scale AI and Stability AI.
Warner said that the responses should be submitted no later than May 26.
