https://www.zdnet.com/innovation/workera-study-ai-skills-upskilling-2026/
AI is no longer emerging in the workplace. It's in the office, and it's shaping how businesses, employees, and their respective interests operate, and it's doing so rapidly.
A new study from Workera found that 80% of companies feel they are more likely to be on track for an AI-enabled future in 2026, up from 67% last year. However, some may be missing the boat on what's most important — their people.
Findings from Workera's 2026 State of Skills Intelligence Report, published Sept. 23, suggest business leaders and working professionals need to find more common ground on how to upskill for an AI world, and Kian Katanforoosh, founder and CEO of Workera, said organizations need to take the lead.
“I would advise any leader right now to think about giving [their] people time to upskill. Don’t think that they will just figure it out without you actually carving out time. With time also comes the psychological safety; give them the psychological safety to experiment [with AI],” Katanforoosh told ZDNET in an exclusive interview.
This year's research surveyed 1,000 salaried professionals working for organizations with 5,000 or more employees in the US. The survey was conducted in July 2026 using the market research tool Pollfish. Year-over-year measurements were obtained by comparing July 2026 results with those from a similar survey conducted with Pollfish in March 2025.
AI skills are imperative, and both companies and employees know it. Workera's survey found that more than two-thirds of employees (67.8%) now use AI tools beyond ChatGPT at least a couple of days a week, up from 39.9% last year. That’s nearly a 30% year-over-year increase in AI tool use.
But that rise in day-to-day use is underscored by a few less-gleaming realities.
Nearly 60% of employees said no time is allocated during work hours for upskilling (56.4%), and almost 43% mentioned a lack of relevant learning materials (42.5%) as the biggest obstacles to improving AI skills. Additionally, employees reported spending minimal time each week on skill development. Over 80% of employees spend just five hours or less per week on training (84.3%).
Katanforoosh said AI upskilling can be cumbersome, with many employees feeling "overwhelmed" with so many ways to learn and little time built into workdays to do it. On top of this issue, the last 12 months have rapidly changed the learning landscape.
"Last year was more about adoption and access, and now we’re past that, and so it’s about outcomes and actions, which creates additional stress," Katanforoosh said.
The 2026 metrics fluctuate minimally from last year's results, indicating that while widespread adoption is increasing, companies are still falling short.
Last year, Workera indicated that most employees hadn't been offered any AI-specific training opportunities. This year, nearly 6 in 10 employees said they've been offered AI-specific training opportunities in the last 12 months
While organizations are providing more training, they are failing to provide proper resourcing and time for employees to use AI in practice.
"People are feeling like they’re cramped; they don’t have the time to learn. On top of that, the pace of innovation has kept accelerating," Katanforoosh said. Still, many employees try to persist.
Just under half of employees (46.5%) said they've used tools not provided by their employer for skill development. Of those respondents, nearly three-quarters (74.6%) said they used fairly mainstream AI tools, including ChatGPT, Claude, and Gemini.
There's an obvious gap here that needs to be closed, and Katanforoosh said it can be as simple as organizations defining what "AI-ready" means for them. Then, a better structure can begin to take shape.
"The second aspect is, if you define the standard, make sure people can measure themselves against that standard ... it turns out if you have a standard and you have a measurement, and then a layer of incentives ... they push people to accelerate their own learning velocity," Katanforoosh said.
Enterprise organizations should minimize barriers to accessing AI tools. However, Katanforoosh acknowledged skills assessments, as Workera offers, can continue to encourage AI upskilling by granting employees more access as their AI skills develop, which also serves as a healthy control.
Organizations must remember that AI is for everyone, Katanforoosh said, and upskilling puts both businesses and employees ahead.
"It turns out, the amount of [AI] slop is probably correlated with people knowing or not knowing how to use AI, and so making sure that you train [employees] ahead ultimately pays off in terms of the [AI] slop that you will see or not see," Katanforoosh said.
Businesses want (and need) AI-skilled employees to succeed, and employees want to feel valued and rewarded for those efforts. The bottom line is that while AI is a great asset, it's only as smart as the people who know how to use it to optimize their tasks and workflows, and employees are well-aware of this gap. Many are betting on people.
Almost 80% (76.6%) of respondents believed they can do their job better than AI. In fact, only 9.5% of respondents said that an AI agent could do even more than half of their job effectively today.
What's more, most employees also believed a human element will remain key to leveraging AI upskilling. Just 7% of respondents said that AI can already evaluate skills better than humans, and nearly two out of five (38.5%) said AI will never surpass humans at assessing skills.
While people still want to be managed by people, Workera's survey indicated that controlled and continuous skill assessment may be a way forward that all parties can benefit from.
New developments like Workera's Ambient, an AI-native skills assessment tool currently being piloted with over 10,000 signups on its waitlist, aim to measure skills in the flow of work.
About two-fifths (41.1%) of respondents said they’d opt into a continuous skills measurement tool like Ambient if they owned the data and controlled what’s shared, while another 31.2% were undecided.
The report detailed that building skills is one of the most effective and affordable ways to grow, though most companies have no way to tell whether their methods are working.
Workera and Katanforoosh are working to change that situation by creating a better route for "learning velocity" with projects like Ambient.
Right now, Katanforoosh said there is a premium for learning velocity, and whoever maintains a high learning velocity has an easier time finding a job.
Historically, Katanforoosh said companies could try to maximize learning velocity by introducing skill measurements to find baselines, and then work to close the gaps as fast as possible.
Workera is aiming to target the "continuous measure," merging learning and work closer together than ever.
"If you manage to merge work and learning, you’re effectively helping someone learn exponentially more than in the past. Today, with AI, we’re able to bring learning closer to work because AI understands unstructured data,” Katanforoosh said.
While privacy and data security remain top priorities for organizations like Workera as they continue to innovate, Katanforoosh said products like Ambient will be a reality sooner or later, and that they’ll be forces for good.
"What tells me that this continuous learning is a good thing is one, it will maximize people’s learning velocity ... But on top of that, I think the consumer and the employee are getting used to getting help from technology."
(Score: 4, Insightful) by Anonymous Coward on Thursday October 01, @06:29AM (2 children)
So you have to "upskill" and "maintain high learning velocity" just to keep up with the Joneses. Basically re-learning your job, but now "with AI!"
And using AI is the bees' knees because it hallucinates and produces non-obviously, subtly wrong output that smarter people than you then have to correct. Or maybe not, because your job was already utter bullshit and nobody was looking at the output anyway. What it doesn't do, is help you learn the actual job. So juniors will grow old but won't grow into seniors. So in a while nobody will be left with actual skill, all they can do is "trust the AI".
And this is a great bright future that you have to work really hard to help make happen, because...?
Somebody enlighten me please.
Oh, I might've already found the answer. This workera outfit sells AI to the office, so a "study" from workera telling you AI is here to stay and camp out in your office, why, that's good solid science of the kind that you just have to trust, no?
A different study [theregister.com] tells us a different thing: AI needs to up its game massively to even just break even. No ulterior motives then, none at all.
(Score: 4, Interesting) by khallow on Thursday October 01, @11:57AM (1 child)
In that light, Workera needs to take care in how it markets its FUD (which it might already do, I'm not interested enough to look). There's good money in damage control from AI misuse - companies will desperately need to appear to try to make things better, if they're caught misusing AI or generate some sort of fiduciary failure through AI. But there's also good money in selling to companies that haven't reached that point - and don't want to appear that they've reached that point. Workera can market itself as the next few steps on the AI advancement ladder. Now that you've purchased this super-great AI tool, here's how you can leverage your workers' performance via training them on this crap. And here's the powerful techniques [penny-arcade.com] you need to get past the Deadline boss.
(Score: 2, Insightful) by Anonymous Coward on Thursday October 01, @04:41PM
So you're saying the real problem is that management has been convinced their company needs to keep up with the Joneses, so they 1) pay to jump on the AI bandwagon, but it fscks things up, so 2) pay to unfsck the fscked things, ah, no, to make a big dog&pony show of how it wasn't AI's fault that AI fscked things up ("damage control"), and 3) pay to seem to've jumped on the AI bandwagon for those that aren't ready yet.
How is this not a scam?
(Score: 3, Insightful) by Anonymous Coward on Thursday October 01, @01:28PM (1 child)
I've noticed that my coworkers are being told (instructed, also ordered) to use Al. In my case, it's helpful when they copy-paste shit with formatting, so I know which blocks of their concerns to ignore. (It' really saved me *days* worth of work. Maybe one time in 5 does Al provide relevance - but in those instances, it's relevance, not instructions, not worthwhile. The correct general-area to look, which is obvious from experience. Those without experience are doing work in an unrelated task, and don't have the permission/experience/job to do the work.)
Members of my team have been straight up instructed to use Al for "everything". For understanding intricate, detailed technical work (for which Al doesn't have context), for which Al produces infuriatingly .... however it creates the infuriation that it does, to the point that the older techs just don't want to touch it (but managers _love_ it! such descriptive explanations). I've asked people that use it to recreate Jira tickets and explain what they want, because I can't figure it out among the 800 words of padding. (When they put in the "prompt" to the LLM that generated the fluff, what they want is clear, and can be cleaned up. No so once it's been ciphered.)
We had a license talk a few days ago. I figured it would grow to be about the cost of an employee - not so! To date, it's only grown to be 1/4 the cost of an employee - from $200/mo to about $1000/mo. I've put in else-wise that the companies need to turn a profit, however, and they're *billions* in debt. I would guess this will double again, and double *again* before things are done. Al is simply pricing 'imself out of the market, slowly. It will become much more untenable than it already is, before things are done. (and - the newest, incoming generation hates it? Hah! can't imagine how that'll impact Al's future..)
So how does it pan out? They've suggested just.. turning off security. Like actually, literally - to the point of, "This requires permissions that you don't have, so if you have the higher-ups install the executable as SUID root, you'll be able to do-the-thing." Happy to say, we laughed at Al and did not do that. How helpful!
sorry, still early in the morning. Hard to keep a set of words from one side to the other. Maybe I should run it through an LLM.
- excessive cost
- using it because ordered to, not because it's helpful
- LLMs, as of weeks ago, suggesting users get others' to disable security and enable root all around
- just leave the vault door open, 'cause people need money, and the vault's got money
- this is at a financial corp.
(Score: 3, Insightful) by turgid on Thursday October 01, @02:06PM
this is at a financial corp.
Oh, do they also have pervasive, continuous surveillance of employees too by any chance?
I refuse to engage in a battle of wits with an unarmed opponent [wikipedia.org].
(Score: 3, Funny) by OrugTor on Thursday October 01, @04:11PM
I don't want to read an article that uses the word 'upskill'. FFS the grammar bot just put a red squiggly line under it.
(Score: 1, Insightful) by Anonymous Coward on Thursday October 01, @06:52PM (1 child)
Customer service is the first to go, and will no longer be staffed by humans. The cafeteria/food court will be a series of talking vending machines. I wish everybody good luck.
(Score: 2, Insightful) by Anonymous Coward on Thursday October 01, @08:23PM
The corporate trend is for featureless boxes you can only make dispense something through an app on your phone, bound to some electronic payment system that tracks your every move.
Better bring your own food and drink to work. Like the old construction lunch box with a thermos in the lid.
(Score: 3, Insightful) by sonamchauhan on Friday October 02, @05:25AM
Put 'paid training' in your performance review process. Lots of companies ask you to put a self development goal.
Locate a short technical training course (typically, delivered in-person over Zoom by some local university). File the course description, price and how the course can help, as a self development goal in the review document. Ask the company to pay for the training and give you a few hours a week to study. Do this every year. In my experience, every 3 or 5 years, a training gets approved.