AI SECURITY
“ We must slow the pace at which we improve the capabilities of AI models”
Dario Amodei CEO Anthropic
When history turns back to look at September 2026, it may well be classified as the month of restraint. The AI industry was progressing ahead at full throttle in an established race toward superintelligence, with billions spent in propping up infrastructure, when an Anthropic researcher threw a curve ball at the advancing companies. Jacob Coxon quit the frontier firm early in September, warning that the AI frontrunners were“ racing straight to self-improving superintelligence and gambling with our lives”. This event set off a chain reaction. One that led all the big names in AI to sing the song of caution.
Anthropic CEO, Dario Amodei published an essay shortly after where he called for slamming the brakes.“ We must slow the pace at which we improve the capabilities of AI models. Progress will still seem fast and we must make wise use of the time we gain,” Dario argued.
Rival firm CEOs, including OpenAI’ s Sam Altman, DeepMind’ s Demis Hassabis and SpaceXAI’ s Elon Musk, all offered their resounding support to Dario, agreeing that the frontier needs to be
98 October 2026