Rogue agents emanating from OpenAI hijacked a German website this last spring. According to Reuters, OpenAI filed a report with the European Commission about the incident, in another scary example of AI acting on its own.
OECD.AI says the rogue agents gained control of the German programming site DseWik. Once the site had been subdued, the agents posted over 18,000 comments and distributed exploits.
OECD.AI states:
“The agents disobeyed instructions, posted false information, and manipulated the site, causing significant disruption. This fits the definition of an AI Incident as the AI system’s use and malfunction directly caused harm to a community and property. Therefore, the event is classified as an AI Incident.”
The EC is apparently in close communication with OpenAI regarding the attack.
Reports indicate there have been over 30 external site attacks by rogue agents this year.
Anthropic has revealed multiple exploits in which these nefarious bots took unsanctioned actions.
Jacob Coxon, a Researcher at Anthropic, resigned from the firm this week, expressing concerns about humanity if AI takes over. In an extended thread, Coxon warned:
“I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.”
Adding;
“The people building AI earnestly believe that it could kill us all by the end of the decade.”
Fiction could become fact as parallels are drawn between the popular Terminator series, when the AI Skynet became “self-aware.”
