Within its new Responsible Innovation initiative, researchers at Penn Engineering discovered that certain features of AI-governed robots carry security vulnerabilities and weaknesses that were previously unidentified and unknown.