• Academics seem to overestimate the degree to which students are using generative AI in their studies
• ChatGPT and other AI tools are preferred by students as productivity tools that cover everything from producing text to improving grammar
• Mixed experience and knowledge of AI among university staff has made it more difficult to judge how students use the technology, offering an opportunity for improvement among academics

A new study has revealed the differences between students and lecturers when it comes to AI use, implying that academics think students are using generative AI too much, whereas students think they’re not using it that much.

Rating their perception of AI use, the academic staff who responded to the research recorded an average of 5.23 out of 7, whereas students recorded 3.11 on average.

The research highlights the differences in how AI is viewed among educators and students, with generative AI the focus. Common uses for the technology include producing written text and checking grammar on existing writing. It is also commonly used for research, finding answers to assignments, and other tasks where finding sources needs to be more efficient.

Understanding AI use

The research from the University of Manchester and Spain’s Universidad del Pais Vasco/Euskal Herriko Unibertsitatea took place between April and June 2024. A school of social science disciplines based at a research-intensive university was chosen for the study, which surveyed 45 academic staff and 86 students.

While the results above indicate that students believe their AI use is low, it may be subject to some underreporting. However, the researchers also add that AI use can be difficult for lecturers to spot. Staff use of AI tools was recorded, with just 14% using a paid AI tool subscription.

Whatever this research tells us about AI use in universities, it should be viewed within context of a two-year old study at a single university.

A complex reality

AI use among students has rocketed over the past few years, with the technology becoming as relevant to coursework and research as Google Docs and Wikipedia. Some educators have resigned themselves to marking work produced with AI, while others have taken more extreme measures.

“There is a tendency for the debate about AI in universities to focus on the risks of cheating or the promise of transforming education, but the reality is more complex," noted Dr Mark Carrigan, Senior Lecturer in Education at The University of Manchester and one of the study’s authors.

Finding a response that universities can apply as a policy for learning requires greater understanding on AI use among students, Dr Carrigan says. “Our findings suggest that staff and students have quite different perceptions of how AI is being used, making it difficult for universities to develop effective responses based on assumptions rather than a clear understanding of students’ experiences.”

The solution, Carrigan adds, is to embrace AI at all levels – for students and lecturers.

“We need to move beyond a climate of suspicion and focus on how these tools can be used appropriately to support learning, while rethinking teaching and assessment for an AI-enabled world.”

• Unsanctioned AI use is increasing, according to a study of Google search data
• 50 terms were identified and their usage analysed, 48% more than two years ago
• The search terms point to use of shadow AI, in which AI tools are employed without corporate oversight or managerial approval

Employees are using AI in the workplace whether its use is approved or not. New research has revealed an increase in “shadow AI” – AI that isn’t really approved for use – and colleagues are relying on the technology to not only assist in work, but also cover their tracks.

The analysis was conducted by consultancy company Acora, which assessed 50 search terms in Google during the 12 months to September 2026. They found the terms were used 161,934 times, 48% more than in 2024 and a 250% increase over ChatGPT’s first year.

Google searches to delete AI chat history, circumvent blocks on AI tools, and to find free AI tools with no signup are among the terms to have increased. While students may be conducting some of these searches – another challenge entirely – there is a growing market for such tools, which could leave organizations looking to better implement AI policies with an even bigger challenge.

Shadow AI

Employing AI tools in the workplace ostensibly offers a productivity boost. The employee should ideally know if they are permitted to use AI, and have attended a session (in person or online) covering the policies and procedures around AI. While this is covered in most organizations, the pace of change with AI has left many businesses playing catch up.

So you have the results, as Acora’s research shows, of increased searches for “unblocked AI” (up 24% in 12 months) and “free AI no sign up” (up 36%) and “AI no sign up” (up 39%) in the same period.

Interestingly, searches for free AI note taking tools have also increased, up 28% on the previous year. If such tools are used in meetings, the information recorded could be stored on remote, unvetted servers, representing a data protection challenge for the organization.

The right controls?

Acora’s study highlights a growing concern: that businesses have AI policies that pay lip-service to the technology’s existence and rules over its use, but ignore the implications of overlooking its risks.

“AI transformation has reached the point where businesses are asking harder questions about what they can trust, how their data is being used and whether the right controls are in place," notes David Rabson, CEO of Acora.

“Concerns around trust, governance, security and data shouldn’t leave organizations stuck between ambition and action.” One solution is to help businesses close such gaps and implement “data foundations, governance and controls in place so they can adopt AI with confidence, rather than putting promising projects on pause.”

• Russian hosting provider Beget was subject to a cyberattack on November 1st, 2026
• While 1000s of websites were hit, cloud storage, VPS hosting, and Kubernetes servers were wiped
• The attack came amidst a spate of drone and online attacks targeting Russian tech companies, presumably by Ukraine-based attackers or anti-Putin elements within Russia

A “massive cyberattack” took Russian web hosting company Beget’s entire operation offline, with abnormal activity recorded immediately ahead of the incident. All levels of its infrastructure were affected, with websites, cloud storage facilities, VPS hosting, and even hosting for Kubernetes containers knocked offline.

The impact of the attack took several days for the company’s experts to assess, with estimates that over 1,000 servers need to be physically checked. This indicates either lasting damage from the attack or infrastructure that was already weakened, perhaps by unpatched exploits.

Beget has been the subject of EU sanctions due to it providing hosting services to authorities in the Russian-occupied Ukraine regions of Crimea, Donetsk, and Luhansk.

Recovering data

Responding to news of the attack, Beget issued a defiant statement, boldly stating “We are not giving up and are continuing to look for ways to recover this data.”

"Restoring all services will take a considerable amount of time, as more than 1,000 servers must be manually inspected.”

The organization told Kommersant (translated from Russian) that most data should be recoverable from backups for the majority of its virtual hosting and cloud infrastructure, adding “We continue to work on data recovery, but due to significant volumes, this process will require a long time.”

However, Beget’s response to the attack is hampered by how it organizes and prioritises data, with many customers not using disaster recovery plans.

Anatoly Sazonov, Head of the Department of Compliance with the requirements of the Softline Group of Companies, told Kommersant, “Beget’s mass failure is an example of a systemic risk that arises from a high business dependency on a single provider. If a website, corporate mail, databases and other services are posted on one site, the provider’s problems can paralyze several areas of business work at once.”

EU sanctions

Beget was founded in 2007 and has become the second largest domain registrar in Russia for .ru domains, overseeing 14.7% of the market alongside providing web hosting, cloud computing, and other services.

As is standard, Beget has encouraged customers to change their passwords and remove and update and access keys, credentials, or other sensitive information that may have been uploaded to its cloud servers.

Beget has already had a difficult year, having been targeted by EU sanctions for providing hosting services in Russian-occupied territories in Ukraine. Whether the domain registrar and web hosting company was targeted because of this, or as a more general piece of cybercrime, remains to be seen.

• AI interviews are discouraging people from fully participating in the job application process
• Jobseekers who are already employed are less likely to accept an AI interview
• 58% of people who attended an AI interview despite reservations would refuse if other jobs were lined up

Automated interviews are becoming a turn-off for jobseekers, particularly if they already have a job, new research has revealed.

A survey from Use.AI of people looking for work in the UK, US, Canada, European Union, and Latin America found nearly a third (29%) of people already in work had withdrawn from a hiring process that involved an AI interview.

Meanwhile, just 14% of those without work withdrew, demonstrating the importance of finding work when unemployed.

Available for consideration

Intriguingly, while the specific figures concerning the 6731 survey subjects differ across all regions, the pattern of a split between those already employed and those who are not is repeated.

Use.AI’s findings indicate that a candidate’s willingness to participate in an AI interview can depend on several factors, including financial security and what other options the jobseeker already has.

Most of the jobs you see advertised today have AI filtering to hopefully determine the best candidates. The businesses going one step beyond this and detaching themselves from the interview process seem to be making a mistake.

“Employers spend considerable effort checking whether their screening systems evaluate candidates fairly. They should be equally interested in who decides to enter those systems in the first place,” said Matthew Freire, Managing Director of Use.AI.

“Offering applicants an alternative route to a human interviewer would allow companies to examine whether the format itself is influencing who remains available for consideration.”

Replacing hiring experts and people who know the job as interviewers may have saved time and money in the short-term. But if no one wants to be hired by a company that relies on AI interviews, then productivity is lost, and along with it, profit.

Automated Recruitment

Searching for a new job increasingly requires the submission of a resume for screening. Those CVs or applications might then find their way to a human, and a call may be scheduled. But many companies are relying on in-house and third party AI interviewers. It isn’t only potential employers, either – some recruitment agencies are staffed by AI interviewers, too.

All of this makes the job application process increasingly challenging, particularly if you’re someone who relies heavily on body language and facial signals. Succeeding in an interview often means building a rapport with the people on the other side of the table or display, and this is something that simply isn’t possible when the person asking the questions isn’t a person at all.

The graphical representation of a programmed LLM may appear smart and attractive, and they have the questions read to ask. But it is ultimately an elaborate piece of form-filling that doesn't demonstrate a person's ability to establish instant rapport and translate that into working relationships.

• Accenture reportedly publishes new research paper littered with what appear to be AI-hallucinations
• The Leapfrogging Development paper has 15 inaccuracies that have been identified so far
• Accenture, which has been demanding its employees embrace AI, has denied that the errors were due to the technology

Top consultancy firm Accenture is under fire once again for publishing papers that include proven AI hallucinations.

The most notable of these, written in collaboration with IE University in Madrid, is titled “Leapfrogging Development: Unlocking Africa’s Potential in Global Value Chains with Artificial Intelligence” and has 15 inaccuracies.

The AI hallucinations were found by AI-detection software publisher GPTZero, which regularly performs analysis on significant new reports. It lists a total of 31 inaccuracies across three papers, with the lion’s share found in that single policy paper.

Spotting the fakes

GPTZero has plenty of previous experience in spotting citation issues with reports from large organizations. A similar scenario developed around auditing giant PwC in August 2026, with GPTZero highlighting their reports as making inaccurate and false claims.

With the Accenture-linked “Leapfrogging Development” report, GPTZero determined 15 AI hallucination inaccuracies, along with a further four flawed citations – that is, citations that don’t exist, exist with a different name, or simply don’t support the report.

“GPTZero contends that hallucinated citations are a clear and present danger to researchers, academics, consultants, students, and anybody else who happens to search the internet for information.”

Previous reviews of reports and research tainted with AI hallucinations have proved that any fake claims that appear in public documents that are published by well known organizations can be recycled and repeated. It’s like an internet version of “Chinese whispers.”

GPTZero adds that “Over time, these recycled hallucinations can ‘poison the informational well,’ even if the original publication is never read by humans.”

In short, AI hallucinations are more dangerous than thought, and Accenture seems prone to publishing them.

Dangerous citations

While the Leapfrogging Development report is the main issue discovered by GPTZero, it is not the only problem. Three documents were reviewed in total.

“Using an automated pipeline and our Hallucination Detector," reported GPTZero, "we surfaced three Accenture products containing a combined 31 hallucinated sources, numerous fabricated claims, and other likely signs of AI generation. To avoid making the same mistake as Accenture, a real human verified each hallucination.”

Speaking to The Telegraph (paywall), an Accenture spokesman claimed the issues surface by GPTZero were down to human error, adding “There is no evidence to support the conclusion that they were generated through AI hallucination. The report is being corrected ahead of a re-release. We will always embrace the opportunity to strengthen the quality of our reports.”

Accenture has recently specified its employees use AI more productively. While these papers are collaborative and produced beyond Accenture’s usual in-house process, they highlight continued AI-related struggles for the firm.

• Adobe for Startups enables new companies founded within the last five years to get its products for free
• Eligible companies will get 12 months of free access to Adobe’s AI-heavy tools
• Firefly Pro Plus, Acrobat Studio, and Semrush are available with the offer

Adobe is launching a new free suite of tools for new businesses, aiming to deliver easy solutions to founders that won’t break their startup budget.

Adobe for Startups will be open to eligible companies founded in the past five years, who will be able to use Firefly Pro Plus, Acrobat Studio including Adobe Express Premium, and Semrush completely free for 12 months.

Marketing the incentive as a way to build branding as early as possible, Adobe is inviting any business founded within the past five years and with under $15 million in funding to apply.

Creation, productivity, marketing

Approved applications to Adobe for Startups will get Adobe Firefly Pro Plus, its all-in-one creative AI studio, which lets you ideate into images, video, and audio. With Acrobat Studio, AI Assistant and Adobe Express Premium are provided for easy content creation and publishing.

Meanwhile, Semrush gives you access to search intelligence and marketing options, illustrating how you business is found online and how it can be optimized.

While program terms and criteria may differ depending on business location, Adobe for Startups offers a useful software platform for growth – and it will help ensure that startups are looking at Adobe's creative AI offering rather than those of competitors like Canva.

• Hackers have shared data concerning Russian casualties in its war with Ukraine on a Russian state website, while also stealing 23TB of data
• Russian state statistics agency Rosstat admitted its servers were targeted over the Halloween weekend
• The organization’s electronic document management system was also hacked and exfiltrated

Russian casualty figures in its war with Ukraine have been kept under wraps since 2024, but have now been displayed on the website of Russian state statistics agency Rosstat following a major breach of security.

Rosstat admitted to a breach over the weekend of October 29 to November 1, 2026, but has yet to admit to any form of exfiltration - despite hacker groups Belarusian CyberPartisans and Silent Crow confirming around 23TB of data was taken.

As well as the casualty figures, the Silent Crow group and its collaborators have boasted of hacking and stealing a document management system, and also claimed to have taken control of vehicles.

Specialized units

The details of what is being considered a successful hack by its perpetrators were published on Telegram. The Silent Crow account explained that “access was gained to a large part of the IT infrastructure of the department, including the central office, regional units and the Main Interregional Center for the Processing and Dissemination of Statistical Information.”

Rosstat reported that its information resources were targeted over the Halloween weekend, says Kommersant (translated from Russian). “The infrastructure worked normally: monitoring systems in a timely manner alerted about suspicious activity, which allowed to quickly start the response.”

The organization’s confirmation added: “Compromising databases and leakage of information protected by law has not been identified.”

Cybersecurity experts, on-site experts, and “specialized units in the field of information security and law enforcement agencies are establishing the vector of the attack and identify the sources of the threat,” the department added.

During its prolonged access to Rosstat's servers, the Cyberguerillas BY and Silent Crow groups revealed human casualties to Russian military personnel since the beginning of the war in Ukraine as 1,584,910.

It also revealed a loss of 587,409 UAVs, demonstrating that this is a classification of combat asset that is in far greater use than tanks, artillery, and manned aircraft.

In supporting its claim of stealing 23TB of data across electronic document management and other databases, the group stated (translated) “The scale of the incident goes far beyond the compromise of individual servers or user accounts. After all, it is about penetration into the infrastructure of the federal department with access to central and regional systems, internal circuits and data sets of public and official value.”

Access to domain controllers of Rosstat unlocked access to the organization’s main data, eventually allowing the hackers to compromise virtualization systems like VMware ESXi, along with infrastructure management systems.

The casualty figures are not likely to be verified by official sources in Russia.

• OpenAI launches new program to protect residents and businesses near its data centers from power outages
• The Grid Protectors Program is designed to prevent OpenAI campuses from “hogging” all the power
• A new development in Effingham County, Georgia, will pioneer the program

OpenAI wants to protect local communities situated around its data centers from power outages, and has launched a new program intended to deliver both reliability and affordability to nearby residents and businesses.

The Grid Protectors Program is launching with the planned Effingham County, Georgia data center campus, which features an agreement with Georgia Power to make power available in a flexible demand scenario, which should cover demand for around 700,000 households.

OpenAI’s future facilities are designed to be more flexible in how they demand power from the grid, and the Grid Protectors Program viewed as a “starting point” for improved reliability for homes, businesses, and of course data campuses.

Grid Protectors pilot

The Grid Protector Program certainly seems a step in the right direction from OpenAI. It comes against a backdrop of growing concerns over the power requirements of AI data centers and protests over their approval, invariably in locations where residents are resistant.

OpenAI says the pilot site at Effingham County incorporates its agreement with Georgia Power to accommodate “up to 1,000 megawatts of flexible demand response.”

This is intended to ensure local residential and business properties have the power available “to help keep air conditioning running in hundreds of thousands of homes during extreme heat. For scale, that is equivalent to the average electricity demand of roughly 700,000 households.”

OpenAI says this is more than enough to cover the combined households of Washington, D.C., and Boston.

Financial assurances

By working to offer communities assurances of flexible energy at its new site, OpenAI clearly hopes that its upcoming data centers are recognized as more suited to locations that have recently come under fire.

Perhaps they want to be a part of those communities? The launch announcement adds that its new compute facilities are more flexible than older facilities, and OpenAI is committing itself under the Grid Protectors Program to collaborate with authorities.

“We will work with utilities, regulators and policymakers to develop practical arrangements for the campuses we build in the future,” it announces, “arrangements that reflect local grid needs, help campuses support the grid when it is under pressure, and pay the costs we create. Long-term contracts and financial assurances can protect other customers if a project is delayed, reduced or canceled.”