Will genAI kill the help desk and other IT jobs?

As AI adoption continues to soar, corporate executives are being forced to make hard decisions on what IT jobs can be automated by the fast-evolving technology and which ones can’t.

It’s a conundrum — especially because experts believe as many as a quarter of IT jobs could be eliminated and replaced by generative artificial intelligence (genAI) tools.

“There have been a lot of layoffs,” said David Foote, chief analyst and research officer with IT research firm Foote Partners. “Companies are identifying people who may have been solid workers in the past, but they don’t fit into the new world driven by the [emerging] economy and the technology they’re making bets on.”

Foote believes from 20% to 25% of tech jobs could eventually be taken by AI.

While AI will reduce or eliminate altogether the need for human input in some areas, it will also enhance productivity, requiring professionals to reskill and adapt to more strategic and creative roles, according to a research note by Foote.

Along those same lines, a survey of CFOs in June by Duke University and the Atlanta and Richmond Federal Reserve banks found that 32% of organizations plan to use AI in the next year to complete tasks once done by humans. And in the first six months of 2024, nearly 60% of companies (and 84% of large companies) said they had deployed software, equipment, or technology to automate tasks previously done by employees, the survey found.

Organizations are using AI to automate a wide range of business process, including paying suppliers, invoicing, procurement, financial reporting, and optimizing facilities utilization, according to Duke University finance professor John Graham, academic director of the survey. “This is on top of companies using ChatGPT to generate creative ideas and to draft job descriptions, contracts, marketing plans, and press releases,” Graham said in the report.

In particular, Foote said, 11 IT-related roles will be affected by AI adoption over the next several years — some in positive ways, others not so positive. (Other industry experts believe that number could be higher.)

The roles likely to be eliminated or heavily automated include: software development, cybersecurity, DevOps, UI/UX design, data administration and management, testing and quality assurance, data scientists and analysts, testing and quality assurance, cloud engineers, technical writers, and IT support and systems administration — including network administration. Ironically AI and machine learning engineering is also becoming more automated through tools like Google’s AutoML, Foote said.

Database administration is also undergoing a seachange as AI-powered systems like autonomous databases (e.g., Oracle Autonomous Database) can self-patch, self-tune, and handle much of the database maintenance that used to require human intervention. Specializing in big data will become more critical for admins.

AI is transforming cybersecurity by automating threat detection, anomaly detection, and incident response. “AI-powered tools can quickly identify unusual behavior, analyze security pattern, scan for vulnerabilities, and even predict cyberattacks, making manual monitoring less necessary,” Foote said. “Security professionals will focus more on developing AI models that can defend against complex threats, especially as cybercriminals begin using AI to attack systems. There will be a demand for experts in AI ethics in cybersecurity, ensuring that AI systems used in security aren’t biased or misused.”

IT support and systems administration positions — especially tier-one and tier-two help desk jobs — are expected to be hit particularly hard with job losses. Those jobs entail basic IT problem resolution and service desk delivery, as well as more in-depth technical support, such as software updates, which can be automated through AI today. The help desk jobs that remain would involve more hands-on skills that cannot be resolved by a phone call or electronic message. “At that point, IT is considering it a unique problem; they’re getting involved in things like code corruption,” Foote said, adding that skills in AI governance and regulation will become more critical as companies face increased scrutiny over the use of AI in security.

Data scientists and analysts, on the other hand, will be in greater demand with AI, but their tasks will shift towards more strategic areas like interpreting AI-generated insights, ensuring ethical use of AI, and working on higher-level model development and validation, according to Foote.

“They will need to focus on building models rather than just analyzing data. This includes ensuring that models are ethical, fair, and explainable, especially when these systems are making decisions in sensitive domains like healthcare or finance. Data scientists will also need deep expertise in the specific industries they work in (e.g., healthcare, finance) to ensure that AI models are aligned with business goals and regulatory requirements,” Foote said.

There will also be a growing demand for data scientists with model selection and optimization tools like AutoML, DataRobot, and H2O.ai which automate much of the machine learning pipeline, from data preparation, process, and analysis to model creation and deployment, according to Foote.

Unlike humans, genAI tools can comb through massive amounts of data quickly, much faster than a technologist could — allowing automated tools to identify problems very easily, according to Jack Gold, principal analyst with J. Gold Associates.

While layoffs among tech firms escalated over the past year, Foote believes companies will begin rethinking their hiring strategies. And that could lead to a hiring sprint over the next several months. “When they embraced automation, they ended up letting people go, but then they decided soft skills and institutional knowledge are important,” Foote said.

That’s because the technology can’t create new product ideas, services, or business strategies; those tasks require critical thinking.

“They thought they could get rid of people, but as it turns out, they need a core of people who understand nuances,” Foote said. “[Organizations] need people who know how to communicate in a collaborative way using verbal and non-verbal skills — particularly people who don’t necessarily have some level of technical skills. These are people who can inspire others and motivate other people.”

Gold agrees: while AI will replace some software developers, mainly those focused on routine or repetitive tasks, humans are still needed to define programs and set parameters. “It means that the software engineer will be more productive and able to write more lines of code, so there might be some reduction in the need to hire software folks. But I don’t see that need fully going away,” he said.

Quality control will still be required, Gold and Foote argued. Gold pointed to the recent CrowdStrike disaster. The security software vendor admitted to using an automated process to push out a bad software update to Windows computers, leading to a cascading crash that affected companies worldwide.

“Remember, AI is only as good as the data set its been trained on,” Gold said.

GenAI will also create new jobs. For example, new tools and machine learning technology must be integrated with traditional enterprise systems — and that will require technologists familiar with both. “Integration is big part of what companies are putting effort into,” Foote said.

In the US and Europe, as many as 300 million jobs are likely to be affected by the arrival of AI, investment bank Goldman Sachs estimated last year. Fully two-thirds of US jobs could be partially automated, with one in four current work tasks completely automated. At the same time, AI is expected to boost global GDP by 7%, Goldman Sachs said in its report.

Career fields with the highest exposure to automation include administrative positions (46%) and tasks in legal professions (44%). Not surprisingly, jobs less likely to be affected tend to be in physically intensive areas such as construction (6%) and maintenance (4%). While automation is expected to disrupt a variety of career fields, including IT, not all jobs will be affected equally. For example, in the legal sector, paralegal jobs are more likely endangered than attorneys — one reason the legal sector’s score is so high, according to Goldman Sachs.

With so many jobs affected globally by AI and machine learning, hiring managers will be on the lookout for candidates with experience in those areas. For example, one recent study indicated programers could more than double the number of projects they complete each week by using AI-assisted code-generation tools.

Those tools are increasingly prevalent in software engineering and, somewhat unexpectedly, have become low-hanging fruit for most organizations experimenting with genAI. Adoption rates are skyrocketing. That’s because even if they only suggest a baseline of code for a new application, automation tools can eliminate hours that otherwise would have been devoted to manual code creation and updates. Tools such as GitHub Copilot, Tabnine, and OpenAI Codex can suggest lines of code, fix bugs, and automate code reviews. This is reducing the need for developers to focus on repetitive coding tasks.

Along those lines, Amazon Web Services CEO Matt Garman recently said that within 24 months “or some amount of time, it’s possible that most developers are not coding.”

While some fear the outright elimination of software developers, others believe genAI automation will allow them and other technologists to be more creative rather than focusing on mundane or repetitive tasks.

Tiago Cardoso, principal product manager at AI-assisted content management company Hyland Software, still believes genAI tools should be used to accelerate programming and coding skills, not replace them. Cardoso pointed to an estimate by the US Bureau of Labor Statistics that predicted jobs for software developers will grow by grow by 25% between 2021 and 2031.

“These figures confirm the demand and need for programming skills, but comfort using AI to support coding will be a skill that employers seek out,” Cardoso said. “Developers should lean into upskilling opportunities and see which AI tools can support tasks like debugging and bug fixing and improving code quality by implementing proactive refactoring so they can focus on sharpening skills that AI systems can’t perform.

“Employers will seek out developers with the ability and openness to adapt to the changing tech market,” he continued. “Developers who embrace these changes and seek out ways to strengthen their skills in pace with innovations in AI will be the most valuable as the role continues to evolve.”

Even with widespread adoption of genAI through 2024, the US economy added more than 158 million jobs, and the tech unemployment rate remained at near historic lows.

“More upbeat assessments contend that AI will augment workers — perhaps the least-skilled ones — rather than replace them. Others argue that replacing workers is harder than it seems because jobs are collections of tasks and AI may not be able to do all of them seamlessly,” Philipp Carlsson-Szlezak, global chief economist with Boston Consulting Group, wrote recently.

Gold also believes estimates about the number of jobs genAI will ultimately eliminate are overblown. A more likely scenario, he said, is that workers will be more productive, and so companies might not need to hire as many people.

“And it may give companies an opportunity to shift workers to more strategic functions,” he said. “I actually don’t think the elimination of jobs, if it occurs in any big [way], will happen for two to three years, as there is still a lot of debugging to go on in making AI programs responsive, effective, and efficient. It’s not as easy as people think it is.”