U.S. House Acts on AI: Microsoft Copilot Use Declared Unauthorized

Key Takeaways

  • The U.S. House bans Microsoft Copilot for staff over cybersecurity concerns, reflecting broader AI regulation efforts.
  • Copilot deemed a cybersecurity hazard by the House's CAO, with plans to evaluate a government-specific version upon release.
  • Microsoft aims to introduce AI tools meeting federal security standards in response to privacy and data breach concerns.

As reported by Axios, the U.S. House of Representatives completely prohibited its staff from using Microsoft Copilot due to cybersecurity concerns.

This action underscores the broader effort by the federal government to balance the internal adoption of AI technologies with the development of appropriate regulations for the sector.

Previously, in June, the House limited the use of ChatGPT to its paid version for staffers, explicitly forbidding the free variant.

Catherine Szpindor, the House’s Chief Administrative Officer, has advised against using Microsoft Copilot within congressional offices. She labeled the application “a cybersecurity hazard.” This decision is based on concerns that Copilot might inadvertently expose House data to cloud services that the House does not sanction.

The directive, as clarified by the CAO’s office to Axios, currently applies only to the commercial version of Copilot. Upon its release, the House will assess the forthcoming government-specific version before making any final decisions regarding its use.

Microsoft Responds To Security Concerns

Microsoft Copilot will consequently be disabled and blocked on all House-operated Windows devices. In response, Microsoft expressed its intention to introduce a suite of AI tools tailored for government use. Those will adhere to stringent federal data security and compliance standards.

Microsoft Copilot, developed in collaboration with ChatGPT creator OpenAI, is designed to function as an AI assistant across various platforms, including as an integrated feature within Microsoft Office applications.

This ban reflects broader concerns similar to those in the corporate sector, where there’s apprehension about using consumer-grade AI tools due to potential data privacy issues.

Companies are increasingly considering or adopting business-grade AI solutions that assure data protection and prevent the information from being utilized in future model training, mitigating the risk of data breaches.

In other news, Microsoft has recently launched Copilot for Security, a solution designed to augment human capabilities and address cybercrime challenges. With it, Microsoft is bringing the power of generative AI to the backyard of security and IT professionals.