Source: Hard AI
Author: Fang Jiayao
On March 29th, according to media reports, the U.S. House of Representatives has strictly prohibited its staff from using the AI tool Copilot developed by Microsoft.
Catherine Szpindor, Chief Administrative Officer of the House of Representatives, stated:
<blockquote>
&#34;The House's cybersecurity department believes that using Microsoft's Copilot AI assistant may lead to sensitive data being leaked to unauthorized cloud services, hence prohibiting congressional staff from using this application.&#34;
</blockquote>
This decision is not an isolated case. Last year, the House restricted the use of ChatGPT by members of Congress, allowing only the paid subscription version to be used while banning the free version. In addition, four U.S. senators proposed a bill last year aimed at preventing AI technology from falsely describing candidates in political advertisements to protect the fairness of federal elections.
Some analysts point out that these measures reflect policymakers' concerns about the risks associated with the adoption of AI technology, as well as government agencies' efforts to ensure data security, privacy protection, and compliance while utilizing AI technology.
In response, a Microsoft spokesperson stated that the company is developing and planning to launch AI tools, such as Copilot, that comply with government security and compliance standards, and intends to release these tools later this year.
The Office of the Chief Administrative Officer (CAO) of the House of Representatives stated that they will evaluate the government version of Copilot upon its release and may reconsider whether to allow the use of this tool, provided that the version meets security and compliance standards

MSFT

The U.S. Congress strengthens control over the use of AI tools to ensure data security and compliance

Is Copilot banned by the US Congress? The House of Representatives is concerned about "leaks"