Updated

Updated · KSL NewsRadio · Jun 25

Hackers Target AI Gatekeepers to Hijack Accounts, Gephardt Warns of Prompt Injection

Updated

Updated · KSL NewsRadio · Jun 25

Hackers Target AI Gatekeepers to Hijack Accounts, Gephardt Warns of Prompt Injection

1 articles · Updated · KSL NewsRadio · Jun 25

Prompt injection is emerging as a new account-takeover risk, with hackers trying to manipulate AI bots that handle identity checks, password resets and other security decisions.
Gephardt said the threat is not only AI fooling people but people fooling AI, potentially talking an automated gatekeeper into bypassing rules meant to protect accounts.
One compromised login can spread further when passwords are reused, turning a single breach into a broader security problem across multiple services.
Strong unique passwords, two-factor authentication, tighter privacy settings and limiting what you post online remain the main defenses as companies rely more on AI-driven account protection.

Sources

As experts admit AI manipulation is unsolvable, are companies creating unacceptable risks by using AI as digital gatekeepers for our personal data?

When an AI is tricked into a costly breach, who is legally responsible: the company, the AI developer, or the customer?

Is the rise of AI in cyber warfare creating an arms race that will soon spiral beyond human control and understanding?

1 total