Nothing to fret about over right here, simply the truth that the folks constructing AI are making ready for it to go badly improper, very quickly. Axios reports that high executives at Anthropic, OpenAI and different AI corporations are privately planning for a public and political backlash after a catastrophic AI occasion. Many insiders count on this to occur inside six to 12 months, probably as a cyberattack that knocks out monetary companies, web entry and even energy and water.
Anthropic CEO Dario Amodei raised the alarm final month, warning {that a} swarm of AI brokers might take over the web, “probably inflicting lots of of billions of {dollars} in harm.” Some safety professionals pushed again. Expel CTO Greg Notch known as the thought “far-fetched.”
Actual examples are taking place, although. A hacker allegedly used Chinese language fashions, together with DeepSeek, to steal knowledge from tens of hundreds of South Korean financial institution clients, then used Anthropic’s Claude Code to search for consumers, in response to CrowdStrike.
The businesses’ planning focuses on educating Congress. OpenAI says its preparedness workout routines cowl many eventualities. “These eventualities usually are not handled as inevitable,” a spokesperson instructed Axios. Anthropic declined to remark.
Nothing to fret about over right here, simply the truth that the folks constructing AI are making ready for it to go badly improper, very quickly. Axios reports that high executives at Anthropic, OpenAI and different AI corporations are privately planning for a public and political backlash after a catastrophic AI occasion. Many insiders count on this to occur inside six to 12 months, probably as a cyberattack that knocks out monetary companies, web entry and even energy and water.
Anthropic CEO Dario Amodei raised the alarm final month, warning {that a} swarm of AI brokers might take over the web, “probably inflicting lots of of billions of {dollars} in harm.” Some safety professionals pushed again. Expel CTO Greg Notch known as the thought “far-fetched.”
Actual examples are taking place, although. A hacker allegedly used Chinese language fashions, together with DeepSeek, to steal knowledge from tens of hundreds of South Korean financial institution clients, then used Anthropic’s Claude Code to search for consumers, in response to CrowdStrike.
The businesses’ planning focuses on educating Congress. OpenAI says its preparedness workout routines cowl many eventualities. “These eventualities usually are not handled as inevitable,” a spokesperson instructed Axios. Anthropic declined to remark.
