OpenAI’s chatbot, ChatGPT, is arguably one of the most significant influences in artificial intelligence (AI). Now, they’re developing an autonomous assistant system known only as an agent capable of performing tasks with the control of a user’s device.
OpenAI’s agent
According to The Information, the agent software will handle web-based tasks such as gathering data, scheduling, shopping, etc. While many companies attempt to incorporate similar systems into their websites, such as Amazon’s Rufus, OpenAI plans to create a universal bot built on an infrastructure similar to their previous products.
With their assistant built on the large language model (LLM) technology, its algorithm is far more sophisticated than that of other competing systems. It uses deep learning techniques and large data sets to understand, summarize, generate, and predict new content.
OpenAI’s agent system’s ability to perform any physical function a human can perform could revolutionize how people use computers.
Autonomous security
A common issue with the bot, however, regards its security.
With this bot’s power and the apparent security threat of giving a corporate AI system unfettered access to private information, it’s essential to understand the global implications of potential cyber threats from an autonomous action agent. And though OpenAI isn’t likely to utilize data maliciously, cyber hacking from outside sources could result in data breaches.
However dangerous the data this new agent holds could prove to be, it can also protect from future threats aimed at systems. The action agent system will likely improve the protection from cyber-threats of previous AI systems and their interest in all industries.
According to Fabrice Deblock with Cyber News, “Industrial cybersecurity experts are seeing their profession transformed under the influence of AI. In their field, artificial intelligence will play a significant protective role through support and assistance to people by enabling more accurate vulnerability detection.”
Though no fully confirmed information exists regarding OpenAI’s system, many consider action agents the next logical step beyond generative AI systems like ChatGPT and Google’s Gemini despite safety concerns.
