Artificial intelligence is not simply helping employees finish their existing work faster. It is quietly encouraging them to take on work that once belonged to someone else. This shift is sometimes referred to as AI workplace task crossover, as AI blurs the lines between traditional job roles and responsibilities.
A marketer can now troubleshoot a website problem. A customer service worker can calculate a refund scenario that previously required help from finance. A small business owner can draft promotional copy, review contract language and examine sales figures without contacting three different specialists.
OpenAI calls this behavior task crossover. The phrase may sound academic, but the shift behind it is already visible inside everyday workplaces. Employees are becoming less restricted by their formal job descriptions and more willing to move between functions when an AI tool can guide them through the unfamiliar parts.
Nearly Half of Specialized AI Tasks Cross Job Boundaries
OpenAI examined more than 800,000 work-related ChatGPT messages from users in the United States for the first report in its new Work at the Frontier research series.
The study found that 16.8% of all work-related messages involved tasks associated with another occupation. Once OpenAI removed generic activities such as writing, scheduling and summarizing, the crossover rate jumped to 43.5%.
In other words, nearly half of the occupation-specific work people brought to ChatGPT did not traditionally belong to their own profession.
That number says something important about workplace AI adoption. Employees are not only asking AI to improve the tasks they already understand. They are using it as a bridge into areas where they have less training, fewer resources or no immediate access to a specialist.
This does not make every worker an expert in finance, legal work or engineering. It does make many employees capable of completing the first useful portion of a task before asking for specialist help.
Sometimes that first portion is enough.
Customer Experience Workers Show the Highest Crossover
The pattern differs sharply between occupations.
OpenAI found that 77% of occupation-specific messages from customer experience employees involved work linked to another profession. Designers followed at 75%, while human resources workers recorded a 69% crossover rate. Legal employees reached 56%, and marketers came in at 53%.
Customer-facing roles naturally encounter problems that do not fit neatly into one department. A support representative may need to understand payment rules, identify a technical issue or explain a contract condition while speaking with a customer.
Previously, that often meant transferring the question, opening another ticket or waiting for a different team to respond. AI gives the employee a way to investigate immediately.
The result is not always full automation. It is fewer handoffs.
That difference matters because handoffs are where many business processes slow down. Every transfer introduces another queue, another explanation and another chance for information to disappear.
Marketing and Engineering Tasks Are Spreading Across Companies
Some occupations absorb work from other departments. Others supply tasks that spread widely across the organization.
Designers, for example, frequently use AI for tasks outside design. Around 35.2% of their messages involved work normally associated with another occupation. Yet design tasks appeared in only 1.7% of messages from workers in other fields.
Engineering showed almost the reverse pattern. Only 18.5% of engineering messages crossed into other occupations, but engineering-related work represented 7.4% of messages sent by non-engineers. Employees across departments were using AI to troubleshoot software, investigate technical problems and work with digital systems.
Marketing moved in both directions.
Marketers used 24.3% of their messages for tasks associated with other occupations. At the same time, marketing work appeared in 8.9% of messages from employees outside the profession, the highest outward share recorded in the study.
That makes sense. Many modern roles now involve some form of marketing, even when the employee does not carry the title. Founders write product announcements. Salespeople create outreach messages. Recruiters draft employer branding content. Customer service teams prepare help articles that also influence retention.
AI has made those small marketing tasks easier to attempt without waiting for the marketing department.
Smaller Businesses Have More Reasons to Use AI This Way
Task crossover appeared more frequently in smaller workplaces.
Among average users, employees in workspaces with two to five seats recorded an outside-occupation task share of 18.9%. That figure fell to 16.3% in workspaces with more than 100 seats.
The gap is not enormous, but the reason behind it is fairly practical.
Large companies can maintain dedicated legal, finance, IT, design and communications teams. Smaller businesses cannot always afford that structure. One employee may already manage social media, customer inquiries, invoicing and basic website updates before AI enters the picture.
Generative AI gives these generalists more room to operate.
A small business owner does not suddenly become a lawyer because ChatGPT helped summarize a contract. Still, the owner may identify unusual clauses, prepare better questions and decide whether professional legal advice is necessary.
That can save time without pretending expertise no longer matters.
Experience Makes Employees More Ambitious With AI
People also appear to expand their AI use as they become more familiar with the technology.
A separate PYMNTS Intelligence study found that 61% of workplace generative AI users had used the technology for at least one year. Among heavy users, that share reached 75%, compared with 42% among light users.
Longer-term users were not simply sending more prompts. They were using AI for a broader mix of activities and more demanding decisions.
Personal users with at least one year of experience completed an average of 11 AI-assisted tasks, compared with 6.6 tasks among newcomers. Experienced users were also more likely to rely on dedicated AI platforms rather than whichever assistant happened to appear inside a phone, browser or search engine.
That progression feels familiar. Many people begin with low-risk requests such as rewriting an email or summarizing a document. Once the tool proves useful, they try planning, research, spreadsheet analysis, troubleshooting and decision support.
The boundaries move gradually. Then suddenly the employee is doing work that would have been forwarded to another department six months earlier.
Job Titles Are Starting to Fall Behind the Work
The research does not prove that AI is eliminating entire professions. OpenAI makes that distinction clear.
A marketer reviewing contract language with AI has not replaced a lawyer. An HR employee troubleshooting a software problem has not become an engineer. A salesperson exploring customer data has not acquired the judgment of an experienced analyst.
But the edges of those professions are changing.
Employees can now complete small but meaningful parts of specialized work without immediately requesting a formal handoff. That could change what companies expect from entry-level workers, managers and small teams.
Job descriptions may eventually include broader responsibilities because AI makes those responsibilities manageable. Companies may also place greater value on employees who can recognize problems, ask useful questions and verify AI-generated output across several domains.
Deep expertise will still matter. So will knowing when a task has moved beyond what an AI-assisted generalist should handle.
The future workplace may not belong entirely to specialists or generalists. It may favor people who can move between both modes.
AI Could Reshape Departments Without Removing Them
The most immediate change may happen inside workflows rather than headcounts.
Finance teams could receive fewer basic calculation requests but more complicated cases. IT departments may spend less time solving simple account problems and more time managing security, infrastructure and AI governance. Legal teams could review better-prepared questions instead of starting every request from scratch.
Specialists remain important. The work reaching them simply becomes more concentrated.
Companies will need to decide where employees can safely use AI to cross traditional boundaries and where firm controls are necessary. Financial decisions, legal interpretations, cybersecurity changes and sensitive HR matters cannot rely on confidence alone.
AI lowers the barrier to attempting unfamiliar work. It does not remove the consequences of getting that work wrong.
The Org Chart Is No Longer the Full Story
For decades, company structures showed where knowledge lived. The marketing team handled campaigns. Finance handled calculations. IT handled technical problems. Employees moved requests through those channels because specialist knowledge was difficult to access elsewhere.
Generative AI has weakened that arrangement.
Knowledge is no longer confined to the person or department that formally owns it. Employees can retrieve explanations, build drafts, test calculations and investigate problems before anyone else becomes involved.
The org chart still shows authority. It may no longer show everything employees are capable of doing.
That is the real significance of AI workplace task crossover. The technology is not merely making individual jobs faster. It is changing the borders between them.
