AI risk is increasing in areas that were previously deemed secure, as these sectors depend significantly on structured, rule-oriented activities. The Wharton data indicating that up to 96 percent of work can be automated mirrors the experiences of employers currently. Significant areas of accounting, financial analysis, data organization, and even standard coding now adhere to recognizable patterns that AI manages extensively. The tasks that are most vulnerable consist of reconciliations, processing invoices, populating tax forms, creating basic financial models, developing standard dashboards, cleaning data, generating SQL, performing QA checks, and scaffolding code. Companies are proactively automating month-end closures, audits, report creation, pipeline training, and code development. Entry-level positions that used to serve as training opportunities are diminishing. What remains justifiable are abilities associated with context, discernment, interdisciplinary communication, regulatory interpretation, and the capacity to connect technical results with business implications. AI continues to face challenges with chaotic real-world uncertainties and establishing trust among different stakeholders. Students can prepare for the future by emphasizing AI oversight, deep knowledge in their field, managing data quality, narrating with data, and gaining experiences that involve clients or teams. In the coming decade, transactional tasks will be the first affected by displacement, whereas higher-value positions will evolve into hybrids that merge domain knowledge with the skills to manage and verify AI systems. - Aditya Nagpal Founder & CEO, Wisemonk Bio: I build and scale global teams for tech and finance firms entering India. I advise HR and hiring leaders on automation, skills strategy, and compliant workforce design, based on years of placing roles across accounting, data, and engineering functions. LinkedIn: https://www.wisemonk.io/
AI didn't suddenly make Accounting, Finance, Data Science, Business, or CS "useless." It exposed how much of their day-to-day work was always pattern-based, rules-driven, and conducted inside digital systems—exactly where AI agents excel. The real pivot is from degree-as-signal to collaboration-as-advantage: the highest value will accrue to professionals who can design, supervise, and collaborate with AI, not compete against it. In every one of these fields, the first wave of automation is targeting "swivel-chair" tasks. Case studies in healthcare show that when agents take over these flows, organizations see double-digit productivity gains and 400-900% ROI, so employers have strong incentives to keep pushing. But this doesn't eliminate humans—it changes what "qualified" looks like. What becomes defensible across these degrees are the "above-the-agent" skills: 1) Framing ambiguous problems and balancing risk, regulation, and human impact. 2) Exercising judgment in edge cases, where a technically correct answer may be the wrong decision for a patient, customer, or community. 3) Relational work: coaching,conflict resolution, and coalition-building across disciplines. 4) Systems thinking: designing workflows where multiple AI agents and humans interact safely, with monitoring, escalation and clear accountability. Students and early-career professionals in these majors can future-proof themselves by explicitly learning to co-produce outcomes with AI. That means moving beyond "using tools" to understanding how AI agents are integrated, where they fail, and how to govern them. Pair a traditional major with: 1) Hands-on experience orchestrating agents across real systems (not just sandbox prompts). 2) Cross-domain literacy (e.g., health + data, finance + ethics, CS + operations) so you can sit where problems are defined. 3) Practice in the "last mile": turning model outputs into decisions, narratives, and changes that humans can understand and act on. Over the next 5-10 years organizations that frame AI as a replacement strategy will hit walls around regulation, trust, and safety. Those that treat AI agents as force multipliers for human teams—freeing clinicians, analysts, and operators from low-value work so they can focus on judgment, relationships, and complex coordination—will see the best outcomes. The real risk isn't choosing the "wrong" degree; it's choosing not to learn how to lead, govern, and collaborate with your non-human colleagues.
I appreciate the opportunity to comment, but I need to be transparent: my expertise is in logistics, supply chain, and e-commerce fulfillment, not higher education or workforce economics. As CEO of Fulfill.com, I've built a 3PL marketplace connecting brands with fulfillment providers, and while I've witnessed firsthand how AI is transforming logistics operations, I'm not qualified to provide expert commentary on college degree valuations or academic career paths. What I can speak to with authority is how AI is reshaping logistics and supply chain roles within our industry. At Fulfill.com, we've seen AI automate inventory forecasting, route optimization, and demand planning tasks that once required dedicated analysts. The data science work around predicting shipping volumes or optimizing warehouse layouts is increasingly handled by algorithms. We've implemented systems that automatically match e-commerce brands with ideal 3PL partners based on hundreds of variables, work that previously required consultants. However, even in our highly automatable industry, the human skills that remain irreplaceable are relationship building, complex problem-solving when shipments go wrong, strategic thinking about supply chain design, and the ability to understand nuanced customer needs that don't fit algorithmic patterns. When a brand is scaling rapidly and needs creative fulfillment solutions, or when unexpected disruptions hit the supply chain, that's where human expertise becomes invaluable. For this particular story about degree valuations and academic career planning, you'd be better served by labor economists, higher education researchers, or workforce development experts who study these trends systematically. I'd be doing your readers a disservice by offering opinions outside my domain of expertise. If you're ever covering logistics automation, supply chain transformation, or how e-commerce companies are adapting their fulfillment operations in an AI-driven economy, I'd be happy to contribute meaningful insights from our work at Fulfill.com.
Rachel Farris CEO, TaxStackAI.com * Managing Partner, Rachel Farris, CPA, P.C. Bio — Rachel Farris is recognized by Forbes as one of America's Top 200 CPAs and by CPA Practice Advisor as a 40 Under 40 leader. She speaks nationally on AI and accounting transformation and has been featured in major conferences, webinars, podcasts, and publications. https://www.taxstackai.com Accounting now faces some of the highest AI-automation risk because so much of the field's foundational work is structured, rules-driven, and repetitive—ideal conditions for automation. Tasks that once trained new accountants are disappearing quickly: bank feeds categorize transactions with little review, reconciliation tools match thousands of items instantly, AI workpaper systems draft audit sections automatically, and tax engines now produce first-pass returns before a junior staffer could assemble the source documents. Even variance analyses, close narratives, and routine client communications—long considered "professional judgment" work—are increasingly AI-generated. The traditional "busy season grind" that justified large entry-level hiring classes has become the easiest target for automation, shrinking the volume of apprenticeship-style tasks that once defined the early career ladder. Despite these risks, the opportunity is significant. As AI absorbs compliance-heavy workloads, the profession is moving toward higher-value human work earlier in the career cycle: advisory, strategic analysis, anomaly detection, risk evaluation, and client-facing judgment calls. These require context, ethics, communication, and the ability to navigate ambiguity—areas where AI still struggles. Many firms are already redesigning roles so junior staff spend less time on manual prep and more time validating AI outputs, interpreting data, and supporting decisions. Over the next decade, the strongest accounting graduates will be those who combine technical skills with the ability to supervise AI systems, evaluate edge cases, communicate insights clearly, and understand regulatory nuance. Accounting is not becoming obsolete; it is becoming amplified. AI removes low-value tasks, allowing accountants to reach advisory-level impact faster than ever before. The real risk isn't majoring in accounting—it's entering the field without preparing to work alongside AI. Those who adapt will find themselves in one of the most leveraged and opportunity-rich careers of the coming decade.
Hey, A major shift from HR perspective is that the being safe based on a degree is no longer going to be determined by degrees that will predominately lead to careers in a highly technical and/or analytical capacity. The new safe degrees will now incorporate soft skills with domain knowledge as well as an ability to adapt, communicate effectively, and lead change. Examples of this shift include Accounting, Finance, Data Science, Business Administration and Computer Science. The jobs that will have a higher risk of being automated by AI will be the highly repetitive jobs that follow predictable patterns i.e. invoice processing, or reconciliation of bank statements and or transaction postings, basic financial modelling, building dashboards, cleaning data, and writing boilerplate computer code, since these types of jobs map directly to the strengths of large language models and the strengths of automation. We are already automating portions of these jobs using AI to draft reports, create code skeletons, find anomalies, create initial performance summaries this work is then being validated, refined and exception-handled by humans. Skills that continue to be valuable, in my experience, are stakeholder management, ethical judgement, prioritization of tasks in an ambiguous environment, the ability to facilitate, and the ability to translate complex and often messy human need into concise problem statements and action recommendations AI continues to struggle with competing objectives, as well as the lack of contextual explanations, and lastly, where there is a need to connect emotionally with others. I've always encouraged the students or those new to their career path to not abandon their undergraduate degree but instead to augment their degree with coursework or experience in the areas of communication, leadership, cross cultural collaboration, and more specifically. Best regards, Ben Mizes CoFounder of Clever Offers URL: https://cleveroffers.com/ LinkedIn: https://www.linkedin.com/in/benmizes/
Full name: Albert Richer Title + org: Founder, WhatAreTheBest.com Bio (2-3 sentences): I run one of the largest product and SaaS comparison platforms online, which constantly evaluates hundreds of tools across accounting, finance, data, and operations workflows. My team tracks how quickly AI is absorbing white-collar tasks, because it directly affects which tools win and which skills stay valuable. That forces us to see, in real time, where degrees are compounding or being hollowed out. Link: https://www.linkedin.com/in/albertricher Traditionally, safe degrees were programs like Accounting, Finance, Business Administration, Data Science, and even Computer Science because they were fundamentally built on repeatable, rules based work. However, these are exactly the kinds of tasks that large models and software agents now perform with increasing efficiency. Invoice coding, reconciliations, basic modeling, KPI dashboards, requirements documents, and even entry level coding are being automated since they follow predictable patterns and have tight feedback loops. The defensible edge is shifting from "I can do the task" to "I can design the system that performs the task, explain it, and earn human trust." Employers will continue automating execution and will keep hiring for judgment, context, cross functional communication, and the ability to translate vague business problems into structured workflows. Students can future proof their careers by pairing these degrees with skills such as problem framing, domain specialization, and leadership. AI will phase out task work. It will reward those who can design, supervise, and manage entire systems.