Read the original article: Steps Organisations Can Take to Counter Adversarial Attacks in AI
“What is becoming clear is that engineers and business leaders incorrectly assume that ubiquitous AI platforms used to build models, such as Keras and TensorFlow, have robustness factored in. They often don’t, so AI systems must be hardened during system development by injecting adversarial AI attacks as part of model training and integrating secure coding practices specific to these attacks.”
The post Steps Organisations Can Take to Counter Adversarial Attacks in AI appeared first on Computer Business Review.
Read the original article: Steps Organisations Can Take to Counter Adversarial Attacks in AI