US Recommendations to Encourage AI Companies to Report Cybersecurity Incidents

Officials in US President Donald Trump's AI task force recommended during a White House meeting on Thursday that government agencies ease their policies regarding legal liability for cybersecurity incidents, aiming to encourage artificial intelligence companies to report serious security breaches.

Bloomberg quoted a US administration official, who requested anonymity because the meeting was not open to the public, as saying that officials believed the standards previously relied upon by agencies to take regulatory action in response to cyber incidents may have become outdated and could deter companies from sharing important information.

The official added that agencies were invited to place limits on legal liability for these incidents for the next 12 months, with the possibility of extending this period further.

According to Bloomberg, the move toward providing incentives for AI companies, rather than threatening penalties, represents a shift from earlier public statements by the administration, which had expressed willingness to rely on existing laws to punish companies developing unsafe products.

Just last month, the same task force warned AI developers that they are obligated to report and address security incidents, or else face unspecified consequences.

Nvidia scrutiny
The meeting, details of which were first reported by Semafor, came amid the fallout from the resignation of Nvidia CEO Jensen Huang following personal relationships he established, including one with a woman suspected of spying for China.

This issue, along with the close relationship between Huang and President Trump and senior administration officials, triggered intense scrutiny, particularly regarding Commerce Secretary Howard Lutnick, who attended the White House meeting.

Trump formed the task force, known as the "Superintelligence Force" (SIF), at the end of September after public concern about the "catastrophic risks of AI reached a level the administration could no longer ignore," following a series of security breaches caused by AI models.

Trump also signed an agreement with leaders of major AI companies, including Nvidia, Anthropic, OpenAI, Alphabet (Google's parent company), and Meta Platforms.

The non-binding agreement endorsed the use of independent auditors and the application of preventive measures in areas such as cybersecurity, biosecurity, and chemical threats, as well as addressing unintended behaviors from AI models. Trump described the agreement as "morally binding."

Days after its signing, the US Federal Trade Commission paved the way for legal actions against AI companies over product safety concerns.

Shared standards and security testing
The official told Bloomberg that Jay Clayton, head of the task force and Director of National Intelligence, indicated during Thursday's meeting that more private companies might join the agreement in the future, adding that he continues to work with signatory companies to establish shared industry standards.

Emil Michael, Deputy Secretary of Defense and Vice Chair of the task force, discussed his efforts with AI companies to develop incident reporting protocols, according to the official.

Also attending the meeting were FTC Chairman Andrew Ferguson, who also serves on the task force; Office of Science and Technology Policy Director Michael Kratsios; and National Cyber Director Sean Cairncross.

Lutnick, Cairncross, and Kratsios also discussed tests conducted on AI models before deployment, based on an executive order issued in June that relies on voluntary participation.

The official stated that the National Security Agency (NSA) and the Center for Advancing Innovation and Standards for Superintelligence, under the Department of Commerce, are responsible for conducting these tests.

Scott Copur, Vice Chair of the task force and Director of the Office of Personnel Management, addressed efforts to protect critical infrastructure and government systems from AI-enabled cyberattacks.

On the same day, Treasury Secretary Scott Bessent said in a speech at a conservative think tank that a cyberattack targeting cooling systems in French refrigerators and a blood bank was supported by Chinese AI models.