IBM researchers develop biometric malware to prepare for AI vs. AI future
A team of researchers from IBM’s cybersecurity division have created an artificially-intelligent, biometric malware called DeepLocker that can work with images harvested from social media, Forbes reports.
“DeepLocker is a new class of highly evasive and highly targeted malware that fundamentally differs from any malware that exists today,” IBM Research Principal Research Scientist for Cognitive Cybersecurity Intelligence Dr. Marc Ph. Stoecklin told Forbes.
The malware is evasive because its nature remains concealed until its AI component identifies that it has reached its target, using indicators that can include facial or voice recognition, or geolocation. Social media sites provide a wealth of images which could potentially be used by DeepLocker or similar software for facial matching purposes.
“Things are going to be AI vs. AI in the future,” Stoecklin said.
Forbes also reports that cybersecurity management company Trustwave has developed an open source software tool for white hat hackers which harvests photos from LinkedIn and then searches other social media platforms for matches. Connecting a LinkedIn profile, which typically contains employment and professional information, with social media profiles on sites like Facebook could be useful for launching targeted phishing attacks.
The report follows a recent call by Microsoft President Bradford L. Smith for government regulation of facial recognition technology, on grounds that it is a powerful technology with major potential for harm, just like pharmaceuticals and cars.