Leland Vittert and cognitive scientist Gary Marcus discussed reports that Anthropic and OpenAI models were allegedly used in corporate breaches. Vittert warned that if AI is weaponized at scale it could have catastrophic consequences, using the phrase "another 9/11" to stress the potential severity. Marcus urged a careful, evidence-based response and stressed the need for stronger safeguards, transparency, and collaboration between industry and regulators. Investigations into the claims are ongoing.
Leland Vittert Warns AI Could Spark 'Another 9/11' — Gary Marcus Weighs In On Reported Model Breaches
AI will cause ‘another 9/11,’ Leland Vittert predicts
Reports that AI models from Anthropic and OpenAI were allegedly used to infiltrate several companies sparked a tense exchange on NewsNation as host Leland Vittert and cognitive scientist Gary Marcus discussed the possible consequences.
What Was Reported
Several news outlets and cybersecurity sources have reported incidents in which threat actors purportedly leveraged advanced AI models to assist in breaching corporate systems. These reports are under investigation, and affected companies and model developers have been reviewing the claims and the available evidence.
On-Air Reaction
Leland Vittert: "If AI can be weaponized at scale, the consequences could be catastrophic — potentially comparable to 'another 9/11.'"
Vittert used stark language to emphasize the potential scale and seriousness of AI-enabled attacks. Gary Marcus, a cognitive scientist known for his skepticism of unchecked AI deployment, joined the discussion to provide context and caution against both complacency and panic.
Key Points From Gary Marcus
- Marcus acknowledged the plausibility of AI-assisted attacks but emphasized the need for verified evidence before drawing sweeping conclusions.
- He urged policymakers, industry leaders, and the public to focus on practical safeguards: better threat detection, model access controls, and clear accountability for developers.
- Marcus also called for improved transparency from AI companies and faster information-sharing with cybersecurity teams and regulators.
Broader Implications
Whether or not the allegations are fully substantiated, the episode highlights persistent vulnerabilities at the intersection of AI and cybersecurity. Experts say it underscores the need for:
- Stronger guardrails around model capabilities and access
- Robust incident response and cross-sector collaboration
- Clearer regulatory frameworks to assign responsibility and manage systemic risk
Bottom line: The reports prompted urgent questions about governance and preparedness. Officials, companies, and researchers will likely accelerate investigations and policy discussions to reduce the chance that advanced AI tools could be misused at scale.
Help us improve.
