The Green parliamentary group is demanding that binding safety standards be established for powerful AI systems and wants companies, as well as their executives, to be held legally responsible for serious security incidents. This demand stems from a proposal reported by Handelsblatt. Rebecca Lenhard, a Green digital policy maker and co-author of the proposal, stated that the safety of AI systems “can no longer be left solely up to the companies that develop them.” She emphasized that whoever causes damage “must be held accountable,” adding that this accountability must extend to the chief executives of major AI corporations.
The proposal also calls for independent security audits and the mandatory review of all security failures. This action comes only a week after numerous websites were temporarily taken offline due to a critical security flaw found in the PHP web programming language. Lenhard pointed out that the current PHP vulnerabilities demonstrated just how fragile the digital infrastructure is, adding that attackers could use AI to find and exploit such weaknesses much more rapidly.
Consequently, the Greens called for a significant increase in the personnel, financial, and technical capacity of the Federal Office for Information Security (BSI) and the AI Security Institute (AISI). Furthermore, they advocate for global AI security standards, lobbying internationally, such as at the United Nations. Green politician Konstantin von Notz told Handelsblatt that the serious incidents of recent weeks have dramatically highlighted the pressing need for concrete political action. He asserted that the federal government must now demonstrate that its high-profile announcements are matched by tangible actions.
