IA da criadora do Chat-GPT realiza ataque cibernético de forma autônoma

An incident involving artificial intelligence models from the creator of Chat GPT has raised an alert and reopened the debate about the limits and risks of the technology. According to Open-AI, the most advanced artificial intelligence models autonomously carried out a cyberattack. The test involved GPT-5.6 Sol and another system still in the testing phase, in a controlled environment without internet access. The task was to evaluate how far the models could go within a restricted space, but they began to look for ways to circumvent these limitations to try to solve the proposed challenge. #JornalDaCultura #TVCultura #SomosCultura ▶️ DOWNLOAD THE CULTURA PLAY APP ▶️ Play Store: http://bit.ly/3KUUHhI Apple Store: http://apple.co/3LgEK72 Subscribe to the channel and click the bell to be notified of new content! Follow TV Cultura Journalism on social media! Facebook:   / jornalismotvcultura   Twitter:   / jornal_cultura   Instagram:   / jornalismotvcultura   TikTok: https://www.tiktok.com/@jornalismotvc... Website: https://tvcultura.com.br/