Artificial Intelligence & Machine Learning
,
Events
,
Next-Generation Technologies & Secure Development
Venable’s Heather West on the Types of Attacks on AI and Securing AI Models
Artificial intelligence is driven by a model, but that model is integrated into everything around the AI. Therefore, we need both AI security and traditional security for any system that uses AI.
See Also: The Role of AI, Machine Learning in Email Security
“We need to be integrating AI into cybersecurity systems,” said Heather West, senior director, cybersecurity and privacy services at Venable. While AI models are “not yet standardized enough at attack,” the rest of the system can be targeted. So following traditional cybersecurity best practices is critical.
That means: “Do the boring work,” according to West. She said conducting risk assessments, determining where to put resources, and deciding which parts of the system need the most protection may not be as “exciting” as what AI can do, but they are needed to protect the entire system.
In this video interview with Information Security Media Group at RSA Conference 2024, West also discussed:
- Attacks on AI systems that target the availability, integrity and privacy of data;
- Highly targeted attacks on predictive and generative AI;
- The impact of regulations and frameworks on securing AI and making it safe to use.
West focuses on data governance, data security, digital identity and privacy in the digital age at Venable. She has been a policy and tech translator, product consultant, and long-term internet strategist, guiding clients through the intersection of emerging technologies, culture, governments and policy. She is a member of the CyberEdBoard.
This post was originally published on 3rd party site mentioned in the title of this site
