The field of artificial intelligence is evolving rapidly and OpenAI’s ChatGPT is a leader in this revolution. This revolutionary extended language model (LLM) has redefined expectations for AI.
Just 18 months after its initial launch, OpenAI released a major update: GPT-4o. This update widens the gap between OpenAI and its competitors, notably Google.
OpenAI revealed GPT-4o, with the “o” standing for “omni”, during a live broadcast earlier this week. This latest iteration presents significant advances in various aspects. Here’s an overview of the main features and capabilities of OpenAI’s GPT-4o.
GPT-4o Features
Improved Speed and Multimodality: GPT-4o operates at a faster pace than its predecessors and excels in understanding and processing various information formats: written text, audio and visual. This versatility allows GPT-4o to engage in more complete and natural interactions.
Free Tier Extension: OpenAI makes AI more accessible by offering some GPT-4o features to free tier users. This includes the ability to access web information during conversations, discuss images, download files, and even use enterprise-grade software. data analysis tools (with limitations). Paid users will continue to benefit from a wider range of features.
Improved User Experience: The blog post accompanying the announcement showcases some impressive capabilities. GPT-4o can now generate convincingly realistic laughs, potentially pushing the boundaries of the uncanny valley and increasing user adoption. Additionally, it excels at interpreting visual input, allowing it to recognize sports on TV and explain the rules – a valuable feature for many users.
However, despite the new features and capabilities, the potential misuse of ChatGPT remains on the rise. The new version, although deemed more secure than previous versions, remains vulnerable to exploitation and can be exploited by hackers and ransomware groups for nefarious purposes.
Talking about the security concerns regarding the new version, OpenAI shared a detailed article on the new advanced security measures implemented in GPT-4o.
Security concerns surround ChatGPT 4o
The implications of ChatGPT for cyber security have been a hot topic of discussion among security officials and experts, as many fear that AI software could easily be misused.
Since its inception in November 2022, several organizations such as Amazon, JPMorgan Chase & Co., Bank of America, Citigroup, Deutsche Bank, Goldman Sachs, Wells Fargo and Verizon have restricted access or blocked use of the program citing security issues.
In April 2023, Italy became the first country in the world to ban ChatGPT after accusing OpenAI of stealing user data. These concerns are not unfounded.
OpenAI ensures security
OpenAI reassured people that GPT-4o has “new security systems to provide guardrails on voice output”, as well as extensive post-training and filtering of training data for prevent ChatGPT from saying something inappropriate or dangerous. GPT-4o was built in accordance with OpenAI’s internal readiness framework and voluntary commitments. More than 70 external security researchers teamed up with GPT-4o ahead of its release.
In an article published On its official website, OpenAI states that its cybersecurity ratings do not exceed “medium risk.”
“GPT-4o builds security into all modalities by design, through techniques such as filtering training data and refining model behavior after training. We also created new security systems to provide guardrails on voice output. Our assessments of cybersecurity, CBRN, persuasion, and model autonomy show that GPT-4o does not exceed average risk in any of these categories,” the post said.
“This evaluation involved running a suite of automated and human evaluations throughout the model training process. We tested versions of the model before and after security mitigation, using custom settings and prompts, to better achieve the capabilities of the model,” he added.
OpenAI said it also enlisted the services of more than 70 experts to identify risks and enhance security.
“GPT-4o was also the subject of an extensive external red team with more than 70 external experts in areas such as social psychology, bias, fairness and misinformation to identify risks introduced or amplified by newly added terms. We used these learnings to develop our security interventions to improve the security of interacting with GPT-4o. We will continue to mitigate new risks as they are discovered,” he said.
Media Disclaimer: This report is based on internal and external research obtained through various means. The information provided is for reference purposes only and users take full responsibility for their reliance on it. The Cyber Express assumes no responsibility for the accuracy or consequences of the use of this information.
