In a nutshell, artificial intelligence is what it sounds like: An artificial system or machine that has human-like intelligence.
AI generally refers to computers and software that perform tasks normally done by humans or that exhibit human-like intelligence to do things such as learn, reason or act. AI can be used to process large amounts of data, identify patterns and provide recommendations.
"The goal of AI," according to a useful primer by Stanford University's IT department, "is to develop machines that can operate independently, adapt to new situations, and improve over time, ultimately enhancing efficiency and productivity across various domains, from healthcare to finance to transportation."
Still, there isn't one agreed upon definition of AI, which sometimes leads to confusion. It also makes regulating AI trickier, since a definition needs to be pinned down to avoid unintended consequences. For example, a broad definition of AI could cover simple spreadsheet software, like Microsoft Excel.
And while "AI" as a talking point exploded recently with the launch of generative AI, the technology has been around for years. In fact, AI technology has already long been part of daily life, used for email spam filtering, language translation, music recommendations, credit card fraud detection, wireless communications and much more.
As AI tools start working and become commonplace, they often fade into the background, becoming just another app or piece of software. So, what's changed recently, and what should you know about it? The most revolutionary update is the availability of generative AI.
What is generative AI?
Generative AI is a specific type of AI that creates content, including text, images, video and audio.
The major tech advance exploded onto the scene in November 2022, when start-up OpenAI released its AI chatbot ChatGPT to the public. Many users were amazed at how the online tool could rapidly churn out a limerick or provide detailed answers to complex questions.
Generative AI can be a general-purpose AI with a wide range of knowledge, contrasted to AI that can do narrow or specific tasks. Users don’t need any technical know-how to get started, which many experts say is a key breakthrough, since it opens the floodgates to rapid adoption.
To use generative AI, users simply type commands in plain English. These so-called prompts give the AI chatbot and other tools all they need to start creating. Don’t like the results at first? You can continuously refine the results with additional notes and questions. Anyone who has used internet search or an online messaging tool can quickly grasp an AI chatbot.
The technology behind generative AI
Generative AI stems from AI systems known as large language models (LLMs), which are trained on massive amounts of data, such as the text of websites, with vast amounts of computer power. These models learn patterns in language to predict the next word. The results are often startling, from in-depth research reports to data crunching, often done in seconds.
LLMs can be trained on images and video, too, to create tools that generate photos, Hollywood-like movies or 3D games. Another top use is debugging and generating computer code, a task many companies are unleashing to their computer programmers.
Tech giants Alphabet (GOOG), Amazon (AMZN), Microsoft (MSFT) and Meta (META) are spending astronomical sums to build huge AI data centers, mostly powered by Nvidia (NVDA) chips, to train and run internal AI models and host AI cloud computing for customers. Many in the tech industry consider generative AI the next big tech platform, a major transition similar to, or perhaps even more important than, the shift from PCs to mobile phones.
Generative AI takes huge amounts of computing power, and electricity, to work. And tech giants are spending so much so fast, some analysts worry about overspending.
Is AI good or bad?
AI, on its own, is neither. It's better to think of it as a powerful tool, like other advanced technologies.
Put into action, AI brings huge benefits and, yes, plenty of risks. Good or bad outcomes come from how humans and society decide to use powerful AI and what guardrails are put in place.
The emergence of generative AI has sparked a new excitement around the potential for good, along with new fears over AI safety.
Consider the potential for AI to boost economic growth, aid scientific discovery and improve public health. Generative AI tools are already helping businesses write emails, summarize documents, create presentations and analyze company data. Small businesses can create slick online video ads without needing any technical expertise. There is a growing list of positive uses of AI in medicine, including sequencing protein structures, analyzing medical images and transcribing doctor visits.
AI apps can help organize family calendars and manage household chores. Or even do the shopping for you, with the emergence of AI agents. There are AI voice assistants, tutors and personal trainers.
Traditional AI brings a range of risks, including the potential for bias or threats related to privacy and security. But generative AI brings at least a dozen new risks unique to the underlying tech. For example, AI chatbots can produce false content, known as hallucinations. These falsehoods can seem authoritative and even come with detailed, but fake, citations.
Generative AI can leak private data, offer dangerous or violent recommendations or infringe on intellectual property. AI chatbots can be tricked into spilling a company's or AI model’s internal secrets, even if they are designed not to. Companies building AI tools are well aware of these risks and trying to find ways to mitigate them, especially for high-risk settings, such as in the military.
Criminals and foreign actors can use generative AI as a low-cost tool to mass produce online spam content or email phishing scams – or for creating propaganda and misinformation in huge volumes cheaply.
In many conversations, the bad aspects of AI often swamp the good. But it’s worth keeping in mind the huge potential of AI to improve society, just as great technological shifts of the past have.
Can I use AI for free?
Yes, plenty of chatbots are free, including tools from OpenAI, Meta, Anthropic and Google. Free chatbots come with limitations, though, such as limits on daily usage. Web search from Google now includes expanded AI answers, while AI web search start-up Perplexity offers free AI results.
You may already use generative AI without knowing it, since the tech is quickly being integrated into popular social media and messaging apps, such as Facebook, Instagram, WhatsApp and X. Meta says it already has 1 billion active AI users, since it's building AI into its apps' search tools. Meta is also releasing a standalone, free AI app.
Consumers are finding AI integrated into their smartphones and PCs, too. Users can get real-time translation of foreign languages during phone calls or text messaging, or have long articles or email chains summarized in a few easy-to-read bullet points.
Paid chatbots and other tools have more features, work faster and can provide better responses, along with fewer or no restrictions on use. More and more businesses are adopting paid tools for workforces to organize email, transcribe meetings, analyze data, create marketing copy, do research and much more.
Pricing varies. OpenAI's ChatGPT has a "Plus" version for $20 per month and a "Pro" version for $200 per month. Anthropic's Claude has plans that run $17 per month and $100 per month. Microsoft's Copilot Pro costs $20 per user per month.
How is AI used in finance?
Like other industries such as medicine and retail, there is a growing list of practical uses of AI in finance. Financial services companies are using it to process customer calls, fight financial fraud, create personalized marketing material, forecast sales, interpret regulations, understand tax law and more.
Bank of America, for example, says 90% of its employees use AI to help with IT issues, health benefits and other employee questions. The company also uses it to help with client questions, summarize research and prepare meeting materials. And its AI customer assistant, Erica, has been used more than 2.5 billion times.
Nearly half of consumers have or are considering using generative AI tools for personal finances, according to a survey by Experian last year. Two-thirds of Gen Zers and millennials are using the tech to manage their finances. Thirty-eight percent of those surveyed said they trusted generative AI as much or more than human advisers. The popular investing app Robinhood will launch an AI virtual assistant this year to subscribers on the Gold $5 per month plan.
You can now export your deep research reports as well-formatted PDFs—complete with tables, images, linked citations, and sources.Just click the share icon and select 'Download as PDF.' It works for both new and past reports. pic.twitter.com/kecIR4tEneMay 12, 2025
Don't be shocked if you find out your personal financial adviser is using generative AI. The CFP Board, a trusted organization for financial planners, published a guide on generative AI earlier this year for Certified Financial Planners.
The guide addresses ethical issues and notes that generative AI can be used for a range of tasks, such as summarizing client meeting notes, conducting initial research and refining marketing materials. But it also highlights that generative AI can produce inaccurate information and that financial advisers are responsible for the final work product.
Because AI chatbots are growing in popularity, it’s no surprise that people are using AI to pick stocks. When it comes to personal finances, be wary of giving away sensitive financial information to AI chatbots and look out for inaccuracies. It still pays to have trusted sources and human judgment.
