Israel cyber chief: We must protect AI-based vehicles from hacking

Every cyber attack on critical infrastructure in 2019 was thwarted

Israel National Cyber Directorate (INCD) chief Yigal Unna at the Cybertech conference in Tel Aviv, 29/01/20 (photo credit: ODED KARNI)
Israel National Cyber Directorate (INCD) chief Yigal Unna at the Cybertech conference in Tel Aviv, 29/01/20
(photo credit: ODED KARNI)
The most immediate technological challenge confronting Israel is to protect artificial intelligence-based vehicles from being hacked, Israel National Cyber Directorate (INCD) chief Yigal Unna said on Wednesday.
Speaking at the Cybertech conference in Tel Aviv, he said that “artificial intelligence is the new battlefield that will accompany us in the near future.”
The INCD chief, who was previously a senior official in the Shin Bet (Israel Security Agency), said that “the immediate challenge before us is artificial intelligence vs artificial intelligence (adversarial AI) – attempts to cause AI-based vehicles to act contrary to their programming in order to cause damage.”
Unna also reported that the INCD and other Israeli agencies have blocked every single cyberattack on critical infrastructure in 2019.
“There have been zero successful cyber attacks on critical national infrastructures in the past year,” said Unna.
Further, Unna presented the cyber incident breakdown that was reported this year to the INCD National CERT.
Out of approximately 8,600 reports and alerts, he said, more than half (4,415) were received from the public and from organizations at the INCD operational 119 center based in Beersheba.
He explained that the other alerts came from detection systems.
Around two thirds (3,233) of the reports from the public and organizations were verified after evaluation as attempted “cyber incidents” – attacks on a variety of computer systems.
Advertisement
Of the reports that were seen as official cyber incidents, around half (48%) were reports about intrusions, and 36% were reported as attempts at information gathering, phishing or other attempts to compromise the potential victim’s system.
The other reports dealt with vulnerabilities in computer systems (7%), malware (5%), availability of service (2%), and authorization (2%).