started · updated
German Green Party demands liability for AI executives
The Green Party parliamentary group in Germany has submitted a motion calling for mandatory security standards for high-performance artificial intelligence systems. The proposal seeks to hold both companies and their executives personally liable for serious security incidents.
Co-author of the motion and digital policy expert Rebecca Lenhard stated that AI security should no longer be left solely to the companies developing the technology. The motion also calls for independent audits, mandatory investigation of security incidents, and the implementation of technical mechanisms to allow for the emergency shutdown of autonomous AI systems.
The initiative follows a critical security vulnerability in the PHP programming language, which highlighted the vulnerability of digital infrastructure. Politicians emphasized that AI could enable attackers to exploit such weaknesses more rapidly. To combat these risks, the Greens are calling for increased personnel, financial, and technical resources for the Federal Office for Information Security (BSI) and the AI Security Institute (AISI).
Entities
Claudia Plattner · Federal Office for Information Security · Green Party · Konstantin von Notz · Rebecca Lenhard · TÜV Association
Claims
What the coverage asserts, and how many sources carry each claim.
- [● 3 SOURCES] 90 percent of affected companies reported highly realistic phishing emails. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] Claudia Plattner is the President of the Federal Office for Information Security (BSI). www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] 52 percent of companies test their AI applications before deployment. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] One in six surveyed companies registered a cyberattack or fraud attempt related to AI. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] Only 11 percent of companies using or planning to use AI have their own procedures for AI security incidents. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] Approximately half of surveyed companies already use AI. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] One in four companies have no rules regarding AI security. www.mannheim24.de · www.op-online.de · www.tz.de
- [● 3 SOURCES] 11 percent of affected companies reported deepfakes involving imitated executive voices. www.mannheim24.de · www.op-online.de · www.tz.de