US official calls for new security designs to address Artificial Intelligence threat

The swift advancement of artificial intelligence (AI) raises significant concerns about potential threats, prompting the need for integrating safeguards into systems right from their inception rather than adding them as an afterthought, emphasized a prominent U.S. official.

Jen Easterly, the director of the U.S. Cybersecurity and Infrastructure Security Agency, expressed unease about the prevailing norm where technology products are released with vulnerabilities, requiring consumers to later address these issues.

Easterly underscored the urgency of adopting a different approach with AI due to its immense power and rapid pace of development. She conveyed these remarks in a telephone interview following discussions in Ottawa with Sami Khoury, the head of Canada's Centre for Cyber Security.

On the same day, agencies from 18 countries, including the United States, endorsed recently developed British guidelines on AI cybersecurity. These guidelines emphasize secure design, development, deployment, and maintenance of AI systems. Khoury emphasized the necessity of considering security throughout the entire lifecycle of AI capabilities.

Earlier in the month, leading AI developers committed to collaborating with governments to conduct testing on new AI models before their release, aiming to better manage the risks associated with the swiftly evolving technology. Easterly expressed satisfaction with the collective efforts made by nations and technology companies to establish technical guidelines for building AI capabilities securely and safely.