Source:

• Forbes, published on September 28th, 2026

By Indrabati Lahiri

Read full article here: Forbes

Source:

• University World News, published on September 24th, 2026

By Riccardo Ocleppo, Founder and Director of OPIT – Open Institute of Technology

For several years, the debate around remote education has been framed too narrowly. We have asked whether online learning can reproduce the experience of the physical university, when perhaps the more useful question is whether it should try to.
 
Students increasingly choose remote education for reasons that go well beyond convenience.
 
It can allow a professional to acquire new skills without leaving a job, a parent to study around family responsibilities, or a student to access a specialised programme that does not exist locally. It can also enable people to learn at different speeds rather than requiring everyone to follow the same timetable.
 
This is particularly important as the shelf life of professional skills shortens. Higher education is no longer necessarily something that happens once, in a particular location and at a particular stage of life. For many students, learning is becoming a recurring activity alongside employment and other responsibilities.
 
But flexibility alone does not make remote education effective.
 
One legacy of the pandemic is that ‘remote learning’ is still sometimes confused with taking a conventional university model and putting lectures on Zoom. High-quality online education requires almost the opposite approach: it needs to be designed for distance from the beginning.
 
The hidden challenges of studying remotely
 
The obvious challenges are familiar: unreliable connectivity, unequal access to devices and the risk of social isolation. But there are subtler problems that institutions also need to address.
 
One is temporal inequality.
 
A university may be geographically accessible to a student while remaining temporally inaccessible. For example, a live session at a convenient European time may take place in the middle of the night elsewhere; or a working student may technically be able to participate in a programme but be unable to attend activities scheduled during working hours.
 
Institutions therefore need to think about accessibility not simply in terms of where students are but also when they are able to learn.
 
There is also the cost of flexibility. Giving students greater control of their schedule is valuable, but it transfers some of the organisational burden traditionally carried by the institution to the learner. Students must decide when to study, what to prioritise and how to remain on track, often while surrounded by the distractions of everyday life.
 
This can particularly affect students who are neurodivergent, those returning to education after a long absence and those balancing study with unpredictable work or caring responsibilities.
 
Another challenge is that problems become less visible online. In a physical classroom, an instructor can often see confusion, frustration or withdrawal. In a remote environment, a student can disengage for days before anyone notices.
 
And then there is AI.
 
Generative AI gives remote students something previous generations never had: almost instantaneous assistance at any time of day. Yet the same technology can remove an essential ingredient of learning: the struggle required to understand something for oneself.
 
The challenge is therefore not simply to give students more technology. It is to design technology that knows when to help, how much to help and sometimes when not to provide the answer.
 
From content delivery to learning architecture
 
The first principle for effective remote education should be to design online-first, rather than asynchronous-only learning.
 
Recorded material, readings and exercises give students flexibility, while synchronous sessions should be reserved for activities where being together genuinely adds value: discussion, problem-solving, debate, mentoring and feedback.
 
At the OPIT – Open Institute of Technology, for example, asynchronous material is combined with weekly live faculty sessions, continuous tutor support and peer interaction. The objective is not to replicate a lecture theatre online but to use each format for what it does best.
 
This also suggests that universities should pay more attention to ‘latency in support’: how long does a student remain stuck before meaningful help arrives?
 
In a campus environment, students have multiple informal opportunities to resolve problems. They can approach a lecturer after class or ask a peer. Those accidental interactions largely disappear online if not addressed properly.
 
Institutions should combine seven-days-a-week online tutors, peer communities, structured office hours and professors’ ‘office hours’ so that a student rarely reaches a dead end.
 
AI should create better questions, not simply faster answers
 
The most interesting development in educational AI is therefore not the generic chatbot. It is the emergence of course-specific systems that understand the curriculum and can scaffold learning. In other words, smart AI tools are designed for specific classes and support students through tough topics instead of just giving them answers.
 
While, at OPIT, the core of the learning experience provided to students is ‘human-led’ (and this will remain core going forward), a well-designed AI tutor can identify what a student is trying to understand, ask questions, provide hints, change the level of an explanation or direct the student back to relevant material.
 
Importantly, it should be grounded in trusted course content rather than treating the open internet as an unquestioned source of truth.
 
There is now encouraging evidence behind this model.
 
A 2025 randomised controlled trial involving undergraduate physics students at Harvard found substantial learning gains from a carefully designed AI tutor. The important qualification is that the researchers did not simply give students access to a general-purpose chatbot: the AI experience incorporated established principles including scaffolding, active learning, targeted feedback and self-pacing.
 
That distinction matters.
 
At OPIT, we have been exploring the same principle through an AI Copilot integrated into the learning environment. Rather than functioning simply as an answer generator, it is grounded in institutional course material and designed to provide contextual support; its behaviour also changes during assessments to protect academic integrity.
 
This is one example of a wider direction higher education should explore: AI as part of the pedagogical architecture, rather than an external tool sitting beside it.
 
UNESCO’s 2025 survey of institutions associated with its Chairs and UNITWIN networks found that nearly two-thirds had already introduced or were developing guidance on AI use, while confidence in its effective pedagogical application remained uneven.
 
The next stage therefore needs to be less about whether universities ‘allow AI’ and more about precisely what kinds of learning behaviour they want AI to encourage.
 
Using data without reducing students to data
 
Online environments produce signals that physical classrooms often cannot: whether students have accessed material, attempted exercises, returned to difficult concepts, or gradually stopped participating.
 
The temptation is to use these signals to predict failure. But prediction is valuable only if somebody acts on it.
 
Recent research on early-warning systems illustrates this point. Machine-learning models were able to identify at-risk students with relatively high accuracy, but the more important element was the intervention that followed: relational, carefully designed communication from instructors prompted students to re-engage with learning activities.
 
The principle should be human-in-the-loop analytics. Data should create an opportunity for a conversation, not an automated judgement about a student.
 
Designing for difference
 
Remote education can also become more inclusive precisely because it does not require every learner to experience education in the same way.
 
Recorded lectures can have transcripts and summaries. Students can pause, replay or accelerate content. Interfaces can offer alternative representations of material. Translation and captioning tools can reduce linguistic barriers. Virtual laboratories and simulations can increasingly give students access to practical experiences without requiring physical proximity to specialist equipment.
 
At OPIT, we have found that features such as recordings, transcripts, flexible pacing and one-to-one support are particularly valuable for neurodivergent learners, but these are not ‘special’ features. They frequently improve the experience for everyone.
 
That is perhaps the larger lesson.
 
The future of remote education will not be determined by whether universities have better video platforms or more sophisticated AI. It will depend on whether they redesign the learning experience around the reality of their students and on whether they can create a genuinely enriching ‘human learning experience’ in an online setting, end-to-end.
 
Universities should measure not only attendance and completion but also the time students spend stuck, the speed and quality of feedback, participation across time zones, opportunities for meaningful interaction and whether technology is increasing or reducing student agency.
 
Remote education is sometimes described as education without proximity. Done properly, it can be something more interesting: education in which proximity, time, pace and support are deliberately redesigned.
 
The goal should be to use the possibilities of the medium to create a form of higher education that could not have existed before.
 
Riccardo Ocleppo is the founder and CEO of Docsity and founder and director of OPIT – Open Institute of Technology.
 

Read full article here: University World News

Artificial intelligence is evolving rapidly and playing an increasingly important role in many industries and aspects of modern life, and it’s important to ensure that this technology is handled carefully and responsibly. With that in mind, the European Union was the first major global power to introduce a comprehensive regulatory framework for artificial intelligence: the Artificial Intelligence Act.

Quick Summary

• The EU’s AI Act entered into force in August 2024, with new provisions rolling out over time.
• It regulates AI systems according to their risk and applies to relevant providers and deployers within its scope across the European Union
• It requires providers and deployers of AI systems to take measures to support the development of AI literacy among their staff and other people who operate or use AI systems on their behalf
• From August 2026, companies also need to be transparent about how and where they use AI.
• Penalties for breaching the AI Act include tiered fines reaching up to 7% of annual turnover.
• Open Institute of Technology’s programs (such as the MSc in Responsible Artificial Intelligence) helps the next generation of tech professionals to understand AI and how to educate others on its responsible usage.

Exploring the European Artificial Intelligence Act

The AI Act was proposed by the European Commission in April 2021, more than a year before ChatGPT was publicly released. The legislation was subsequently negotiated and amended as AI technologies, including generative AI and chatbots, developed, before the European Parliament adopted it in March 2024. It was approved by the EU Council a couple of months later, and came into force on August 1, 2024.

However, its rules and requirements did not all become applicable right away. Instead, provisions of the AI Act have been rolling out gradually, with the rules for high-risk AI systems in certain sensitive areas, including employment, applying from December 2027.

This has led to some misconceptions about the Act being “postponed” or “suspended” somehow. In actuality, many of its provisions are already in effect, and companies across the European Union – as well as those who do business in Europe – need to be aware of which rules they currently need to follow, as well as the schedule and deadlines for upcoming provisions.

Initial Regulations

Since February 2025, Article 4 of the AI Act has required providers and deployers of AI systems to take measures to support the development of AI literacy among their staff and other people who operate or use AI systems on their behalf. This doesn’t necessarily mean that they need to have formal training or some sort of qualification in AI, but organizations should take measures to support the development of AI literacy appropriate to their staff, the AI systems they use, and the context in which those systems are used.

Companies using AI chatbots to help them prepare commercial quotations, for example, should know how to do this in a safe and responsible manner. Similarly, businesses that use AI to generate content or summarize documents should also have relevant educational systems and policies in place to ensure that this is done correctly.

To comply with this part of the Act, companies are encouraged to:

• Set clear internal guidelines about the use of AI.
• Host AI tool training sessions for workers.
• Have procedures in place to verify AI-generated content.

What Changed in August 2026

On August 2, 2026, several new AI Act provisions became applicable, including transparency obligations for certain AI systems under Article 50. To be more precise, companies are now obliged to make it clear to their users, customers, and clients when, where, and how AI is being used; i.e., if a customer is chatting with an AI bot, they need to be informed of that right from the start.

This also covers AI-generated or AI-manipulated content. For certain AI-generated or AI-manipulated content, providers must use machine-readable markings to enable its detection, while deployers have separate obligations to disclose certain deepfakes and AI-generated or manipulated text published on matters of public interest. The Act doesn’t say that companies can’t use generative AI, but they do need to be open about it.

The regulation is particularly strict in situations when AI has been used to entirely or partly produce content that is designed to inform the public about important subjects, such as:

• Safety
• Energy
• Health
• Consumer rights
• Taxation

Again, companies aren’t banned from using AI to create this kind of content, but deployers generally must clearly disclose AI-generated or manipulated text published on matters of public interest unless it has undergone human review or editorial control.

Looking Ahead

From December 2, 2027, additional provisions of the AI Act will apply to certain high-risk AI systems, including systems used in employment and recruitment, such as those used for CV screening, candidate analysis, or other recruitment decisions. These may fall into the “high-risk” category, mentioned earlier, and even though these rules don’t apply today, experts encourage companies to get ready for the next regulatory phase by:

• Remaining aware and up-to-date with the AI Act’s existing provisions and rollout schedule.
• Auditing their usage of AI systems to understand their current and future obligations.
• Establishing detailed guidelines, policies, and clear workflows around AI usage.
• Training employees on how to use AI correctly and responsibly, in accordance with the Act.

Companies also need to be aware that violations of the AI Act can lead to significant penalties, with fines for certain infringements reaching up to €35 million ($39 million) or 7% of a company’s total worldwide annual turnover, whichever is higher.

Key Takeaways

• The EU’s AI Act takes a risk-based approach to artificial intelligence, with different requirements applying to AI systems depending on their risks and uses.
• At this time, providers and deployers are required to take measures to support AI literacy among relevant staff, while certain AI systems and uses are subject to transparency requirements
• In the coming years, businesses that use AI for “high-risk” activities, like recruitment, will need to ensure they comply with emerging provisions.

Become an AI Leader at OPIT

The AI Act proves that AI literacy is becoming an increasingly important skill to master, and the next generation of tech professionals will need to be able to not only understand AI, but educate others on how to use it responsibly. You can take your first steps on that journey by enrolling in the Open Institute of Technology’s MSc in Responsible Artificial Intelligence.

Technology is evolving at an increasingly rapid rate, and two major innovations are being explored together as quantum computing meets artificial intelligence. This was the topic of discussion for the first episode of the Open Institute of Technology’s “OPIT EDGE” podcast, which saw host Levi Mandji sit down with AI expert Matteo Zangani to talk about what comes next.

Quick Summary

• The first episode of the OPIT EDGE podcast focused on the implications of quantum AI.
• Quantum AI is the combination of AI and quantum computing principles.
• It will allow AI models to process data and broaden their knowledge at unprecedented speeds.
• Expert Matteo Zangani believes this technology could present new opportunities for humans.
• He believes it can be used to automate tedious tasks, freeing people to become more creative.
• OPIT’s programs (BSc in Digital Business, MSc in Responsible AI and MSc in Applied Data Science and AI) can help students expand their knowledge on AI technologies.

Introducing OPIT EDGE

OPIT EDGE is the student-driven podcast project from the Open Institute of Technology. Each episode is oriented around a specific sector of the technological world, from Web3 to software engineering, and episode one was entitled “Let’s Talk QUANTUM AI and AI STRATEGY with an Industry Expert.”

The “industry expert” in question was Matteo Zangani, CEO and tech lead at Alpha Quantaris, an innovative startup specializing in the integration of AI and quantum computing for applications in fields such as finance, logistics, and energy. Zangani is also the Manager of Artificial Intelligence and Automation at Flutter Entertainment and has experience as a Senior Data Scientist at Stellantis.

The Age of Quantum AI

At this point, AI needs no introduction. This exciting and immensely powerful technology has quickly transformed life as we know it, seeping into numerous industries and aspects of the world, automating workflows, fast-tracking processes that were once tedious and slow, and opening up countless new opportunities in fields as diverse as education, healthcare, scientific research, and marketing.

The concept of “quantum AI,” however, is still relatively little-known. In a nutshell, this term refers to the integration of two emerging technologies: AI and quantum computing. Quantum computing is an advanced type of computer processing that utilizes the laws of quantum mechanics to solve certain problems at an extremely rapid rate, potentially faster than conventional computers.

When quantum computing power meets AI agents, it’s believed that this could bring about a dramatic evolution in AI capability. Models could potentially perform certain learning and optimization tasks faster than before, which could lead to breakthroughs in fields such as:

• Drug discovery
• Financial forecasting
• Cybersecurity
• Logistics/supply chains

The State of Play

For now, quantum AI is more of a theoretical concept than an active one. The technology is still in the research and development phase and isn’t being used on a daily basis, with firms like Alpha Quantaris still exploring how it might work and the advantages it could bring. Quantum processors already exist, although large-scale, fault-tolerant quantum computers suitable for broad commercial use remain under development.

The Positive Implications of This Technology

Despite the fact that true, large-scale quantum AI isn’t yet here, experts are already imagining the positive changes it could bring to the world when it finally arrives. Some of its many suggested advantages include:

• Faster and Cheaper AI Training: In applications where quantum computing provides an advantage, quantum AI could potentially operate faster than conventional approaches. Smaller amounts of energy will be needed to level up agents’ intelligence levels and make breakthroughs.
• Higher Accuracy: Quantum-informed AI models should be able to deliver more precise responses compared to conventional neural networks, and won’t need to use as much memory or computational power to do so.
• Scientific and Social Breakthroughs: As quantum AI develops, it’s believed that it could contribute to breakthroughs in fields like climate science and medical research, potentially helping researchers better understand diseases and address complex challenges.

The Expert’s View

Despite the potential benefits of quantum AI, some fear that this technology could hasten the rate at which AI infiltrates industries and takes over jobs that are traditionally filled by humans. There are concerns about how humanity will cope in terms of making sure people can remain employed and earn a living when AI agents will be able to do almost everything more quickly and efficiently.

Zangani argues that this viewpoint is overly negative. He believes that humanity has already undergone civilizational shifts throughout the ages as new technologies have emerged and taken over tasks, but also opened up new opportunities, and he feels that the rise and evolution of quantum AI will play out similarly:

“If we go back two centuries, more than 90% of the people worked in agriculture, and now it’s less than 2%. But we still have a lot of jobs. So we need to embrace transformation and be open. I don’t think AI is going to replace anyone’s work, and I don’t think it’s hype at the same time. It’s a transformation period.”

He goes on to state that even in its infancy, quantum AI is already unlocking exciting opportunities, and it’s up to humanity to embrace those chances moving forward. He also believes that this type of technology will be able to automate repetitive and tedious tasks, freeing humans to focus more on creative pursuits and higher value uses of their time and energy.

Zangani’s Advice

When queried on how aspiring students and AI leaders of tomorrow can prepare for the age of quantum AI, Zangani shared his view that success in AI and tech in general isn’t just about technical skills. He notes that students will naturally need to learn skills such as Python programming and statistics, but also highlights the value of soft skills, such as curiosity, teamwork, and having a growth mindset.

For those who are just starting in this field, he suggests building projects and experimenting with AI in their free time, honing their skills and putting their passion for technology to good use. Even if some of these projects fail, he believes that this is all part of the process towards becoming a more competent professional in later life.

Key Takeaways

• Quantum AI is an emerging technology with vast potential for scientific and social breakthroughs.
• While there are concerns about this technology affecting jobs and employment, experts believe it will also unlock new opportunities.
• Mastering key AI-related skills today will help students and graduates alike be better prepared when the technology becomes more widespread.

Get Ready for Quantum AI with OPIT

If you’re excited about the potential applications of quantum AI, or eager to expand your understanding of AI technology in general, the Open Institute of Technology can be the perfect point of entry. OPIT offers numerous courses oriented around AI and digital technologies, from our BSc in Digital Business to our MSc courses in Responsible AI and Applied Data Science and AI. Check out the OPIT site to learn more and apply today.

Artificial Intelligence (AI) is reshaping the cybersecurity landscape, and not just for the better. While AI-powered security solutions are proving effective at detecting and preventing cyberattacks and malicious programs, agentic AI is also being embraced by bad actors and hacking groups to make their attacks potentially more effective and dangerous.

Quick Summary

• AI technology is taking an increasingly active role in cyber-attacks today.
• It can identify targets, bypass protections, harvest credentials, and steal data autonomously.
• Attacks are likely to become more sophisticated and complex as AI gets smarter.
• The coming years will see swarm attacks, deepfake-as-a-service technology, and other new threats.
• Those wishing to help entities defend themselves can consider an MSc in Cybersecurity at OPIT.

The Age of Autonomous Cybercrime

Autonomous cybercrime isn’t a theory or a prospective threat on the horizon; it’s an active danger, threatening businesses, governments, and individuals across the globe today. Where cyberattacks may have once involved months of planning and manual work, some attacks can now be planned and executed much faster, with significantly less human involvement.

AI can carry out reconnaissance to identify and map targets autonomously, potentially much faster than human attackers working manually. It can also be used to bypass some defensive systems, steal login details, and automate parts of an attack before conventional defenses can start to respond.

This danger is becoming more widespread, with a Signicat report finding a 2,137% increase in deepfake fraud attempts over the previous three years. The report also found that 42.5% of detected fraud attempts in the financial and payments sector involved AI.

A New Era of Cybersecurity

Before AI emerged, cybercriminals regularly developed new tools and techniques to deceive their victims, steal data, and cause chaos in cyberspace. But almost all of their attacks were manually driven. They had to do most of the hard work and heavy lifting, which effectively limited the speed, efficiency, and overall risk level of their attacks.

Now that AI has entered the equation, experts feel that the “Manual Phase” of cybersecurity is all but over, with two new phases emerging from 2022 onwards:

• The Generative AI Phase: Beginning around 2022, this was when attackers increasingly experimented with generative AI technologies to streamline and enhance their campaigns.
• The Autonomous Phase: From 2025 onwards, the autonomous phase involves AI not merely helping attackers, but effectively replacing them. The latest agentic models can do far more than write scam messages or generate malware; they can orchestrate complex, multi-stage attacks, at scale, while human attackers sit back and wait for the results.

The PRAR Loop

Many analysts fear that this new phase of cybersecurity could quickly spiral out of control because of AI’s propensity to become smarter and more effective at carrying out its duties as time goes by. They cite the “PRAR Loop” as the mechanism that cybersecurity businesses and at-risk entities need to be aware of from this point on:

• Perceive: The AI reads and understands its environment.
• Reason: It relies on its own training and data access to plan its next step.
• Act: It puts its plans into action, rapidly and without need for human oversight or instruction.
• Reflect: It reviews the impact of its actions and uses the results to adjust subsequent actions or plans.

This is what makes AI-powered cyber-attacks so concerning. Unlike traditional malware, which generally requires attackers to modify or redeploy it, AI-powered systems can adapt their behavior and automate iterative tasks without requiring a human to direct every step.

The Five Stages of an AI Attack

While individual attacks can vary widely in their planning and execution, a typical deepfake strike might include the following five stages:

1. Reconnaissance: This is the point at which the AI identifies its target and learns as much about it as possible, often using publicly accessible data, like company websites, social media profiles, and press releases.
2. Construction: Again relying on publicly available data, such as videos and audio files, the AI can construct a realistic likeness of a figure of authority within the business, such as the Chief Financial Officer (CFO).
3. Approach: The deepfake is deployed, either in video or audio form. The attacker uses the AI-generated imitation of a real business figure to demand that a lower-level employee take some sort of action, like transferring money or sharing files.
4. Verification: This is the stage at which attacks may be thwarted, but it depends on whether an employee takes the time to verify the deepfake figure’s identity. If they don’t, the attack will almost invariably succeed. If they do, they may realize what’s happening and take action to stop it.
5. Result: Again, depending on what happened in the previous stage, the attack may either succeed or fail. If it succeeds, data, funds, and credentials may all be lost or stolen.

Looking Ahead

By 2027, some experts and forecasts anticipate that AI-powered cybercrime could evolve significantly, potentially enabling malicious groups to deploy coordinated, multi-agent attacks and assisting attackers in identifying vulnerabilities and developing exploits.

At the same time, there are people on the other side of the equation using AI to build better defenses and guard against these kinds of attacks, and working to level the playing field.

These “guardians of tomorrow” are already learning the tools of their trade today on courses like OPIT’s  Master’s Degree (MSc) in Enterprise Cybersecurity, which provides graduates with the technical and managerial expertise they need to develop new security solutions and lead bold cybersecurity initiatives.

Key Takeaways

• AI-powered cyber-attacks are likely to get more dangerous in the years ahead.
• Current attacks often revolve around deepfakes or AI-generated exploits, but new methods are likely to emerge.
• It will be up to cybersecurity students, graduates, and experts to lead the counter-charge and develop AI-powered defenses to stop these attacks before they do damage.

Learn How to Safeguard the Systems of Tomorrow at OPIT

If you’re concerned about the rapid rise of AI-powered cybercrime and would like to play a part in stopping it, you might like to consider signing up for the MSc in Enterprise Cybersecurity at OPIT. It will give you the tools you need to build the stronger, smarter defenses that businesses are going to need to protect themselves from AI threats for years to come. Visit the OPIT site to learn more or fill out an application.

Source:

• Forbes, published on August 11th, 2026

By Gary Drenik

In a difficult business environment, artificial intelligence can look like an obvious answer for executives seeking faster work, lower costs and new efficiencies. But as AI moves into everyday business use, the more important question is whether those gains will last.

New research led by Professor Jan-Emmanuel De Neve of Saïd Business School, University of Oxford suggests companies may be making a strategic mistake if they treat AI mainly to cut headcount. The research argues that businesses focused too heavily on automation risk weakening the human capabilities they will need for growth, including creativity, judgment, institutional knowledge and leadership development.

“It’s a classic manifestation of status quo bias,” says De Neve. “It’s far easier for an executive to look at a spreadsheet and imagine using AI to streamline what people already do than it is to do the heavy strategic lifting of reimagining how technology can produce entirely new value. Treating AI as a headcount-slashing tool might yield quick, visible cost savings, but it is ultimately a path of organisational contraction that trades away a company’s long-term capacity to innovate.”

The distinction is central to AI strategy. Automation replaces tasks people currently perform. Augmentation expands what people can do, giving employees more capacity for problem solving, customer relationships and higher-value decisions.

A recent survey from my company Prosper Insights & Analytics shows why this matters. In its survey of U.S. adults, Executives and Business Owners were more likely than Employees to say they already use generative AI, at 53% compared with 38%. That suggests leaders may be closer to the strategic promise of AI, while employees are closer to the disruption it creates in daily work.

Prosper Insights & Analytics also found that 31% of employees are concerned AI will cause job losses, while 11% are concerned they personally will lose their job because of AI. A further 25% of employees say they do not trust that AI has their best interests in mind.

Those figures matter because AI adoption is also a trust exercise. If employees believe AI is being introduced to make them easier to replace, they are less likely to experiment with it openly, challenge its outputs or show managers where it can improve the business.

Professor Ashish Kumar from Trinity Business School calls this “an agency and incentive problem.” Senior leaders often see AI through the lens of strategy, efficiency and long-term opportunity, while employees deal with the practical consequences when new tools disrupt existing workflows. Kumar argues that workers are more likely to embrace AI when they feel involved in the decision, understand how it will help them, and trust that the technology is being introduced to support their work rather than threaten it.

De Neve argues that augmentation creates a different relationship between workers and AI. “When you choose augmentation, you unlock a compounding growth cycle. Employees engage with curiosity and agency, becoming ‘pilots’ of the technology rather than passive passengers.”

The danger of an automation-first approach is especially clear when companies reduce junior hiring or cut entry-level roles. These roles are often viewed as easy to automate because they involve repeatable tasks, basic analysis or administrative work. Yet they are also where future leaders learn judgment, build relationships and understand clients.

“When you cut human compensation and junior roles to fund technology, you hollow out your internal leadership pipeline and destroy the psychological safety required to innovate,” says De Neve. “The AI revolution will not be won by the organisations that replace people the fastest, but by those that empower them the best.”

The issue is not whether companies should ignore efficiency, it is whether efficiency becomes the whole strategy. Professor Guillaume Coqueret of Emlyon business school argues that AI adoption depends on both individual adaptability and organisational factors including leadership, incentives, culture and the choice of tools. Legacy organisations face a harder challenge because AI often has to be introduced into older processes, established habits and uneven levels of confidence.

That alignment also affects quality. One emerging risk is “workslop,” AI-generated content that looks useful but lacks the substance needed to move work forward. Alan Lerner, professor at the Open Institute of Technology – OPIT, describes workslop as “AI generated work content that masquerades as good work, but lacks the substance to meaningfully advance a given task.”

The danger is that AI increases output volume while shifting the burden onto employees who must check and interpret unclear work, creating the impression of productivity while increasing the effort needed to make AI output reliable.

Prosper Insights & Analytics data points to similar concerns. Among employees surveyed, 37% say AI needs human oversight, while 34% are concerned it can provide wrong information or hallucinations. These responses suggest many employees are not rejecting AI. They understand that its value depends on human judgment.

That judgment is harder to sustain when employees are anxious, overloaded or unsure how their work is being evaluated. Debora Nozza, Assistant Professor at Bocconi University, says AI can reduce some tasks while increasing the mental burden around others.

“In theory, AI should reduce the burden of work. In practice, it often changes the nature of that burden,” says Nozza.

Prosper Insights & Analytics also found that 17% of employees say AI makes them anxious. That does not mean AI adoption should slow, but companies need to pay closer attention to morale, training, communication and workload. If AI leaves employees less able to apply their judgment, the expected efficiency gains may prove smaller than leaders hoped.

The most successful AI strategies will likely be judged by more than adoption rates or cost savings. Companies will need to ask whether AI is improving work quality, helping employees make better decisions, strengthening retention and building the next generation of leaders.

Automation can make old processes faster. Augmentation can help companies develop new sources of value. Companies that recognise this may be better placed to turn AI from a short-term efficiency tool into a lasting source of competitive advantage.

Read full article here: Forbes

Source:

• The AI Journal, published on March 18th, 2026

Source:

• EFMD Global, published on April 10th, 2026