What is Generative AI?

Artificial intelligence (AI) is a vast branch of computer science concerned with a development in software that allows computer systems to perform tasks that imitate human cognitive intelligence, such as visual perception, speech recognition, decision-making, and language translation. Generative AI is an artificial intelligence technology that uses machine learning algorithms to generate content. This can be in the form of text, art, video, images, and more.  

‍

Generative AI works by processing large amounts of data to find patterns and determine the best possible response to generate as an output. The AI is fed immense amounts of data so that it can develop an understanding of patterns and correlations within the data.

Its understanding works by utilizing neural networks, making it capable of generating new outputs for users. A neural network is a mathematical model used in machine learning where each “neuron” in a neural network receives input signals, performs a computation on them using a weighted sum, and applies an activation function to produce an output. Neural networks are trained on large data sets, usually labeled data, building knowledge so that it can begin to make accurate assumptions based on new data.  A popular type of neural network used for generative AI is large language models (LLM).

Machine learning

Machine learning is a subset of artificial intelligence (AI) that involves developing algorithms and statistical models enabling computer systems to learn and program themselves from experiences without being explicitly programmed. In other words, machine learning involves creating computer systems that can learn and improve on their own by analyzing data and identifying patterns, rather than being programmed to perform a specific task. Machine learning has several subsets including generative AI.

Generative AI

Generative AI is a type of AI that is capable of creating new and original content, such as images, videos, or text. This is achieved through the use of deep neural networks that can learn from large datasets and generate new content that is similar to the data it has learned from. Examples of generative AI include GANs (Generative Adversarial Networks) and Variational Autoencoders (VAEs).

Natural Language Processing (NLP)

This is a field of AI that focuses on understanding, manipulating, and processing human language that is spoken and written. NLP algorithms can be used to analyze and respond to customer queries, translate between languages, and generate human-like text or speech. This form of AI is not made for generating new outputs like generative AI does but more so concerned with understanding.  

Generative AI

‍Generative AI and NLP are similar in that they both have the capacity to understand human text and produce readable outputs. The two differentiate in that generative AI uses generative adversarial networks (GANs) which is an approach to generative modeling that uses deep learning methods to autonomously learn patterns in input data and create outputs. The main concern with generative AI is to produce content.  

Large language models (LLM)

‍Large language models are supervised learning algorithms that combines the learning from two or more models. GPT and ChatGPT fall under the LLM scope. This form of AI is a machine learning model that is trained on large data sets to make more accurate decisions than if trained from a single algorithm. An example of LLM is OpenAI’s GPT-4.

‍

ChatGPT

‍(Generative Pre-trained Transformer) A product of OpenAI, is a generative AI system that uses natural language processing, including a large language model called GPT-3 to understand and generate human-like text, answer questions and more. ChatGPT is considered generative AI because it can generate new text outputs based on prompts it is given.

Bing AI

‍Bing AI is an artificial intelligence technology embedded in Bing’s search engine. Microsoft implemented this so that users would see more accurate search results when searching on the internet.  

DALL-E 2‍

his is a text-to-image generator developed by OpenAI that generates images or art based on descriptions or inputs from users.

‍

Individuals can use generative AI like ChatGPT similar to a search engine or chatbot. If you ask the AI model a question it will produce a response based on guidelines provided. For example, we asked ChatGPT the following:

What is cyber security? In 50 words.

“Cybersecurity refers to the protection of computer systems, networks, and digital data from unauthorized access, damage, or disruption. It involves implementing measures and practices to safeguard against cyber threats, such as hacking, malware, phishing, and other forms of cyber attacks, ensuring the confidentiality, integrity, and availability of information.” -Chat GPT

ChatGPT allows you to set parameters and prompts to assist the AI in providing a response, making it useful for anyone seeking to discover information about a specific topic.

Quality

A generative AI model will not always match the quality of an experienced human writer or artist/designer. Their outputs are limited to the data that was given to them to process. For example, ChatGPT was given data from the internet up until September 2021 and might have outdated or biased information. It is possible that in some cases generative AI produces information that sounds correct but when looked at with trained eyes is not.

Context

These models do not appropriately understand context and rhetorical situations that might deeply influence the nature of a piece of writing. While you can set parameters and specific outputs for the AI to give you more accurate results the content may not always be aligned with the user’s goals.

Ethics

Because Generative AI technology like ChatGPT is trained off data from the internet, there are concerns with plagiarism. Its function is not so simple as asking it a question or giving it a task and copy pasting its answer as the solution to all your problems. Generative AI is meant to support human production by providing useful and timely insight in a conversational manner. Similarly, Generative AI is susceptible to IP and copyright issues as well as bias/discriminatory outputs.  

Privacy

Generative AI can learn from your prompts, storing information entered and using it to train datasets. With that data in the system, it is possible that if someone enters the right prompt, the AI could potentially use your company’s data in response to a query. If the information you entered contains sensitive files or data such as intellectual property or know-how, financial reports, confidential internal documents, or sales numbers, it could become part of the third-party AI model and potentially available to others, creating privacy, intellectual property, and security risks if the appropriate guardrails are not in place.

With the capability to help people and businesses work efficiently, generative AI tools are immensely powerful. However, there is the risk that they could be inadvertently misused if not managed or monitored correctly.  

Darktrace can help security teams defend against cyber attacks that use generative AI. Darktrace provides risk and compliance models for Darktrace DETECT™ and RESPOND™ that make it easier for customers to put guardrails in place to monitor, and when necessary, respond to activity and connections to generative AI and LLM tools such as AutoGPT, ChatGPT, Stable Diffusion, Claude, and more.  

Darktrace’s Self-Learning AI makes it possible to detect generative AI activity that may deviate from company policies or best practices, bringing AI to each customer’s unique data, and it learns the day-to-day workings of every user, asset, and device – building an understanding of your business’s unique ‘pattern of life’.  It can detect even subtle anomalies that could indicate a threat to your business and autonomously respond, containing the threat in seconds.