Recently Udacity, a division of Accenture, announced a deal to offer a Master of Science in Artificial Intelligence that it says could be completed for as little as $3,500.
Udacity aims to address a skills gap in AI learning. Although 90% of workers use AI on their jobs, three in four drop AI tools in the middle of tasks because they lack skills, systems and trust, Udacity reported.
Meanwhile, in April learning company Pearson released a Value of IT Certification Candidate Report in which the percentage of candidates that intended to earn a certification in AI and machine learning rose from 17% in 2022 to 35% in 2024.
Udacity’s graduate program in AI is accredited through Woolf, a partner of the Saïd Business School at the University of Oxford and Harvard Business Publishing. It follows Accenture’s $1 billion investment in LearnVantage, a comprehensive learning and training service focused on helping tech professionals develop skills to grow the AI economy.
The curriculum consists of 12 Udacity nanodegree programs in addition to a capstone project, and students log approximately 2,250 hours. Nanodegree programs are a fast track for tech professionals to get experience in a certain skill set. Udacity has seven nanodegree programs as part of the core curriculum and 40 as electives, says Jared Molton, vice president of consumer at Udacity. The graduate AI program is backwards compatible, meaning students can apply nanodegree work toward the master’s.
Each nanodegree program includes real projects such as building a chatbot. Mentors grade the projects and provide feedback.
Although the courses are asynchronous, students have the choice to set up meetings with mentors.
Instructors develop the courses, but they’re not involved in teaching them on a day-to-day basis, according to Molton.
“We hire them to develop the course, build it out, make sure it's right,” Molton says. “Every year we'll update it, but we don't have any type of expectation for continued engagement with our learners. It hasn't been part of our model.”
The degree also includes a capstone project in which students demonstrate their expertise in the areas of the program.
Molton says tech professionals can get admitted into the Udacity AI graduate program without a bachelor’s degree through performance-based admission.
A degree is included in a Udacity subscription for $249 per month for 13 months, plus a one time $199 enrollment fee, Molton says.
If you take the master’s degree program full time, you could complete it in 13 months, but the company says 18 to 24 months is more typical. Under 24 months would be just under $5,000. Students who completed other Udacity programs can get credit and start further ahead in the Udacity master’s degree program.
During the Udacity program, students create a portfolio in GitHub with the relevant code for projects they work on in the program, according to Molton.
The program offers coursework in data science, software development and building out AI systems, as well as multiagent support.
Companies such as Amazon Web Services, Google, Meta and Microsoft contribute to Udacity’s curriculum for AI courses.
Do Tech Professionals Need a Degree in AI?
Graduate programs in areas like AI prepare tech professionals for long-term innovation, while certificates are “tactical” and keep people current on skills, according to Neil Sahota, AI adviser for the United Nations and cofounder of the UN’s AI for Good initiative.
Graduate degrees offer foundations in areas such as optimization, reasoning, decision theory, human-AI systems design, AI governance and ethics, Sahota said.
“These don’t change every six months,” Sahota said. “They shape someone into a long-term innovator, not just a tool user. For organizations that need people who can evaluate risk, architect systems, or lead AI transformation, that depth matters.”
Meanwhile, certificate programs allow tech professionals to stay current in areas such as model APIs, agent frameworks, vector databases, evaluation tooling and prompt patterns, Sahota says.
“Certificates signal that a professional can navigate the practical reality of today’s systems,” Sahota says.
“The real power comes when people combine both,” Sahota says. “A degree gives them the ‘why.’ A certificate gives them the ‘how right now.’ In AI, the people who can bridge those two worlds are the ones leading major deployments inside industries like finance, healthcare, and logistics.”
Sahota notes that a graduate degree in AI will need to stay current more than that of other tech skills because AI is still fluid, he said.
Tech professionals in demand include those who understand mortgages, supply chains, or clinical workflows and can build or understand AI for these areas, Sahota says.
Learning how to learn AI involves speed.
“AI tools evolve faster than curricula,” Sahota says. “Techniques like chain-of-thought prompting, retrieval-augmented generation, or agentic workflows were obscure a year ago and now drive enterprise systems.”
Learning AI skills also involves “problem-framing” rather than “model-tuning,” Sahota says.
“People who know how to translate messy business problems into AI tasks outperform those who only know how to call an API,” Sahota says.
Sahota sums up the best way to learn AI:
“The professionals who thrive in AI are the ones who continually refine how they learn, question, and adapt, and who can turn AI into a force multiplier for real-world impact,” Sahota says.
Mary Hall, a professor and director of the Kahlert School of Computing at the University of Utah, notes that AI graduate degrees that are useful should have mathematical foundations.
“An AI graduate degree is potentially useful, but must include rigorous mathematical foundations to understand the technology that powers AI advances,” Hall says. “Like computing, AI will continue to advance, but will build on these mathematical foundations.”
She says that whether it’s a degree or courses, machine learning skills will be key.
The Kahlert School of Computing offers an AI track for graduate students as well as a new AI minor for undergrads. The minor covers automation of tasks that need human-like intelligence, including learning, perception, reasoning and decision-making.
A Shift in AI Learning
Sahota sees programs such as Udacity’s graduate degree as part of a shift in higher education.
“The real story is the unbundling of the traditional university model,” Sahota says. “By partnering with Woolf for accreditation and Harvard Business Publishing for content rigor, Udacity is signaling that the future of advanced education will be modular, distributed, and competency-driven.”
AI training, however, must keep up with the fast-moving developments in the tech industry, he suggests.
“My concern is durability,” Sahota says. “AI skills today have a half-life of 6 to 18 months.” He adds that the strongest AI education programs will resemble software. That means “continuously updated, version-controlled, and driven by real-world use cases from industry.”
He adds, “Udacity is moving in the right direction, but the real differentiator will be whether their program can keep pace with the agentic, multimodal, and autonomous systems that will define AI work in 2026 and beyond.”
Udacity’s Molton sees AI skills as pivotal to success in business and tech.
“Skills-based hiring is real and it's here to stay, and our core offering is based on helping learners achieve the skills that they want so they can be successful in a role that they want to either be in or grow into,” Molton says. “That's not going anywhere.”