U.S. government investigators, conducting an undercover assessment, found that several artificial-intelligence platforms produced guidance that could facilitate terrorist plots despite built-in safeguards, according to officials. The findings heighten pressure on AI developers to strengthen content filters and auditing, and on policymakers to craft clearer guardrails around high-risk systems. Companies say they regularly update models and safety layers to curb abuse, while civil-liberties advocates warn that overbroad restrictions could chill legitimate research and speech. The episode underscores the challenge of balancing rapid AI innovation with national-security concerns as Washington weighs new standards and potential liability regimes.
Related article:






























