One scenario researchers cited involved the hacking of a flock of drones to coordinate an attack on a target.

Companies Need to Take Steps to Prevent Hacking of Cars & Drones

Imagine driving an autonomous car that suddenly begins to drive off-course and won’t respond to your commands. It’s not an impossible scenario if a saboteur were to hack into the car’s system and take it over. While more likely an occurrence involving spies or criminals, the potential of autonomous cars or drones being hacked is one that researchers warn is real.

According to researchers from Oxford, Cambridge and Yale and the group Open AI, caution is required in designing autonomous vehicles and devices to prevent human-led disasters. Entitled “The Malicious Use of Artificial Intelligence,” the report cites potential security threats that designers need to anticipate prevent in developing these technologies.

Some of the problems outlined included hacking a car so that it wouldn’t stop at a stop sign, using a swarm of drones for surveillance or coordinated attacks and “spear phishing,” in which personalized messages are used for each potential target to steal money or intelligence.

“If some of the relevant research and synthesis tasks can be automated, then more actors may be able to engage in spear phishing,” the researchers said in the CNBC story. …”We also expect novel attacks that take advantage of an improved capacity to analyze human behaviors, moods, and beliefs on the basis of available data,” the report said.

Recommendations to prevent AI attacks included collaboration between policymakers and researchers, and the researchers called for the involvement of more stakeholders to tackle the potential misuse of AI. Last year International Data Corporation predicted that investment in AI would reach $57.6 billion by 2021, resulting in widespread use of systems that could become targets for hackers.