Feds accuse China of ‘systematic’ distillation of U.S. AI models
A joint advisory alleges Chinese companies are using sophisticated systems to route millions of data requests to US AI models across different accounts and platforms.
A joint advisory alleges Chinese companies are using sophisticated systems to route millions of data requests to US AI models across different accounts and platforms.
Static checklists and annual penetration tests leave agencies with dangerous blind spots. True resilience requires moving from reactive attestation to continuous, automated validation.
July’s intrusions reached water controllers that sat on a cellular link no city network scan would find. Naming an owner and paying for the fix are decisions a utility can make this fiscal year, out of money it already applies for.
SonicWall customers have confronted a barrage of attacks for years, including five actively exploited vulnerabilities in SMA 1000 appliances since late 2025.
The nations warn that governments and industry can no longer treat quantum codebreaking as a distant or theoretical possibility.
WatchTowr founder and CEO Ben Harris
Mandating ‘kill switches’ for AI agents would threaten the security of America’s critical infrastructure and undercut U.S. AI leadership. Congress must reject the AI Kill Switch Act.
NEA partner Aaron Jacobson
The communications product company disclosed 22 total Wednesday, all but one of which was rated “critical” at 9.0 or higher.
Agency red-teamers got initial access to both organizations they tested, but one quickly isolated and shut down the attempts from going further.