Mayorkas Unveils DHS Strategy to Secure AI

The first meeting of the Department of Homeland Security‘s AI Advisory Board was held on Monday this week. The board aims to improve its handling of artificial intelligence technology deployment and defend the technology against external threats.

Tackling AI biases and security risks

During a Q&A session with the reporters at the RSA conference and a keynote speech with the employees, Mayorkas discussed AI security and the board’s first meeting discussion results. Members were speaking about the issue of AI security in essential facilities. They also touched on the human rights implications of AI implementation in DHS tasks, with a few highlighting its potential to be used as an extension of the existing implicit bias.

AI systems working in colorism and sexism are a couple of long-standing problems experts and the federal government keep addressing because the thought leaders and the government are minorities who don’t bother testing the products to see whether the systems work with women and people of color. In addition, the White House and Justice Department also began projects directed at doing away with biases brought by the AI systems.

The DHS board, which will meet quarterly, was disclosed. It includes people like OpenAI CEO Sam Altman, government officials active in AI, and owners of critical infrastructure and technology brewers. Mayorkas stated that there are more than 20 AI uses, from focusing on hurricane damage insurance claims to recently helping the Federal Emergency Management Agency (FEMA) train refugee case agents.

Global cooperation against AI threats

DHS officials have highlighted in their reports that they created a portal listing where AI technology has been deployed or rescinded due to disparate concerns among various stakeholders. The DHS officials stressed that the department is looking to expand its workforce in AI-related areas, and 3000 applications have already been received for the job roles in this field.

Mayorkas shared that the board initiated the process of drawing up outlines of strategy and plans for preventing possible threats to AI and reducing the risks of undesirable consequences during AI application. The Supplementary work on Board was already satisfactory. The fact that this is the makeup of international AI leaders is also critical because it will work on legislative harming with other countries.

However, defending AI systems will not be achieved solely by having global players on board. Such a scenario requires the partnership to build a solid defense mechanism.