The Computer Science department will be at the European Researchers' Night 2026 / Vísindavaka 2026 26. September 2026, 12:00-17:00 at Laugardalshöll in Reykjavik.

The Computer Science department of University of Iceland has a couple of booths there, showcasing our activities in a way accessible for the general public. Come and see to experience our interactive exhibitions:

Our booths at the previous European Researchers' Nights:

• European Researchers' Night 2025 / Vísindavaka 2025
• European Researchers' Night 2024 / Vísindavaka 2024
• European Researchers' Night 2023 / Vísindavaka 2023
• European Researchers' Night 2022 / Vísindavaka 2022
• No 2020 nor 2021 European Researchers' Nights due to COVID pandemic
• European Researchers' Night 2019 / Vísindavaka 2019

This is just a small selection, the Computer Science department has even more exhibitions!

---

Parts of this event are in the context of our cybersecurity activities and the ECCC/EU co-funded projects ICEDEF – Defend Iceland and Eyvör – the National Cybersecurity Coordination Centre of Iceland (NCC-IS).

The University of Iceland invites applications for a full-time position as Assistant Professor in computer science or software engineering within the Department of Computer Science in the School of Engineering and Natural Sciences. In accordance with the University of Iceland's regulations, the duties of academic staff are divided into three main areas: research, teaching, and administration.

Field of work

• Teaching within the department, including introductory courses
• Supervision of master's and doctoral students
• Conducting research in computer science, software engineering or related fields
• Attracting domestic and international research funding
• Participation in the administration of the department and teaching development

Qualification requirements

• Ph.D. in computer science, software engineering or a closely related field
• An active research record with publications in international venues in the field of computer science or software engineering
• Good command of English, both written and spoken
• Experience in university teaching and student supervision is an advantage
• Experience in attracting research funding is an advantage
• Experience in interdisciplinary collaboration and/or of collaboration with industry is an advantage
• Good collaborative and interpersonal skills, with the ability to communicate knowledge effectively
• The ability to teach in Icelandic is an advantage. Appointees are required to attain proficiency at the B1 level in Icelandic within four years of appointment. The University of Iceland provides support to help staff achieve the necessary language skills.

Applications are particularly encouraged from candidates working in one or more of the following areas, although applications from other areas of computer science are also welcome:

• Human-computer interaction. User-centred software engineering, requirements engineering, user experience and interaction design, accessible and inclusive design, together with qualitative and quantitative methods for user research.
• Applied artificial intelligence. Software development supported by generative AI, based on formal and informal specifications; verification and validation of AI-generated code; together with the design, orchestration and evaluation of AI agents that carry out tasks with limited supervision.
• Cyber-physical systems. Embedded systems, edge computing, robotics, and distributed or resource-constrained systems.
• Cybersecurity. Security and resilience of software, systems and networks; cryptography and secure protocols; vulnerability analysis; monitoring of and response to security incidents; together with the human dimensions of cybersecurity.

Application process

The position will be filled for a term of five years, with the possibility of a permanent appointment thereafter, cf. Paragraph 3, Article 31 of the Rules of the University of Iceland No. 569/2009. Applicants who meet the relevant requirements may apply for expedited promotion immediately following recruitment under Article 3 of Regulation no. 1300/2020 on promotion of academic staff at the University of Iceland.
The processing of applications, evaluation of candidates` qualifications, and appointment to the position shall follow the provisions of Act No. 85/2008 on Public Universities and the Rules of the University of Iceland No. 569/2009.

Applicants must submit the following documents with their application:

1. A cover letter outlining, among other things, how the applicant meets the qualification requirements
2. Diploma(s) or, where applicable, a certificate of academic record
3. Academic curriculum vitae (CV)
4. List of publications
5. A report on research work and other professional work undertaken
6. A statement of research and teaching plans if appointed
7. Information on potential referees who may be contacted

The application must indicate which of the applicant`s publications, up to eight in total, are considered most significant with regard to the position in question. Applicants should submit only these publications with their application or provide information on where they are electronically accessible. If a submitted work has multiple authors, applicants must clarify their individual contribution to the work. Application materials that cannot be submitted electronically should be sent in two copies to the Science and Innovation Department, University of Iceland, Sæmundargata 2, 101 Reykjavík.

All applications will be answered, and applicants will be informed of the outcome once a decision has been made. Applications may remain valid for six months after the application deadline.

In hiring for positions at the University of Iceland, the University`s equality policy is taken into account. Attention is also drawn to the University of Iceland`s language policy.

Since 1911, the University of Iceland has been a leading institution for business and societal progress in Iceland. It is a forward-looking research university and an active participant in international scientific collaboration.

The University of Iceland is the largest university in the country, offering nearly 400 programs at the undergraduate, master`s, and doctoral levels. Approximately 14,000 students are enrolled, including over 2,000 international students.

More than 1,800 staff members work at the University of Iceland. The university emphasizes attracting a diverse community of employees and students to an international academic environment, where academic freedom, professionalism, equality, diversity, and inclusion are prioritized.

Application deadline

Application deadline is 02.10.2026, 23:59 UTC+0.

For further information contact

Kristján Jónasson

jonasson@hi.is

Hafsteinn Einarsson

hafsteinne@hi.is

Online application form Note: you can switch to English and while it offers you to sign in with an Icelandic Id number or even electronic Id, non-Icelandic residents can also simply register a new user.

The 17th EAI International Conference on Digital Forensics & Cyber Crime (ICDF2C 2026) was successfully held in Reykjavik.

Thanks to everyone who made this event possible!

Proceedings have been published and can be accessed via SpringerLink and you should be able to get even 20% off:

ICDF2C 2027 will be in New York. Submission deadline is 15 December 2026.

This is my first AI-generated blog post (but it is manually reviewed and corrected, so this is no AI slop): after debugging for hours an Ansible script for updating (using fwupdmgr and apt) a Spark DGX that includes a reboot, I involved AI to solve my problem.

Before the automation, I finished the manual firmware and kernel update with a manual

/sbin/reboot now

While this manual approach worked, my automated Ansible-based approach that I tried then, did freeze the network connection and while I could still log-in via keyboard and mouse, the network devices where not visible in system logs. Only pulling the power plug helped to bring the networking back.

As an LLM helped me debugging, that LLM had all the context needed to create this blog post:

When running updates via Ansible on the newer NVIDIA DGX Spark platforms, executing a standard ansible.builtin.reboot module immediately following a major firmware and package upgrade can cause the target node to completely drop off the network. The system will stop responding to pings.

If you check via the physical console (HDMI and USB mouse and keyboard luckily still working) the logs after a ansible.builtin.reboot, you will find a quiet, catastrophic platform trace:

acpi NVDA8800:00: platform device creation failed: -16 (EBUSY)

The primary PCIe Host Bridge Controller crashes, rendering the OS completely blind to PCI devices such as the network. As a result, physical network interfaces (like the Realtek enP7s7) vanish entirely from the system — running lspci returns absolutely nothing.

While Spark DGX forums catch the downstream symptoms of this fragility — for example, citing PCI DOE mailbox timeouts during warm resets — they usually miss the exact software trigger and offer no solution for Ansible-based reboot automation. Here is why it happens and how to fix your Ansible playbooks permanently.

---

The Root Cause: The SSH Handshake Trap

The standard ansible.builtin.reboot module operates on a synchronous protocol. It opens an SSH connection, triggers the reboot, and then clamps that TCP socket open while polling the machine's status to watch it go down and verify when it returns online.

When systemd attempts to dismantle the network driver stacks while at the same time, Ansible is actively fighting to hold that SSH socket open, a severe timing race condition triggers on the unified NVLink-C2C/PCIe switch fabric. Because a warm, i.e. software reboot keeps the motherboard power sub-planes fully energized, the EBUSY (-16) error flags stick in the hardware registers, locking out the PCIe bridges on the subsequent boot cycle.

---

The Solution: Staging and Fire-and-Forget Automation

To fix this platform trait, we have to implement two distinct changes in the Ansible script:

• Assuming that the script involved firmware and package updates: Rearrange the execution order: Stage firmware capsules first using fwupdmgr, run apt package upgrades second, and consolidate the layout into a single reboot at the absolute end to minimize fabric disruptions.
• Decouple the reboot execution: Force a low-level kernel storage flush, request an immediate hardware reset vector jump, and instruct Ansible to drop its hands off the connection before the network stack can stutter.

By wrapping the reboot command in async: 10 and poll: 0, Ansible passes the instruction string to the kernel and instantly terminates its own SSH pipeline that it established from your remote administration machine from where you started the Ansible script. The fabric remains completely quiet, allowing the system to cycle cleanly.

Here is the hardened, production-ready playbook snippet:

    # ==========================================
    # CONSOLIDATED DECOUPLED REBOOT
    # ==========================================
    # We use fire-and-forget (async/poll) to drop the SSH connection 
    # BEFORE driver stacks cycle, preventing the PCIe fabric race condition.
    - name: Data-safe sync and single hardware reset
      ansible.builtin.shell: "sync && /sbin/reboot now"
      async: 10
      poll: 0
      when:
        - reboot_required_file.stat.exists or firmware_staged | bool
        - not ansible_check_mode

    # Explicitly hold execution to accommodate firmware flashing and NVLink link-training cycles
    - name: Wait for hardware initialization and fabric training to complete
      ansible.builtin.wait_for_connection:
        delay: 60
        timeout: 1200
      when:
        - reboot_required_file.stat.exists or firmware_staged | bool
        - not ansible_check_mode

---

The Takeaway

The above wait_for_connection task will patiently pause automation, i.e. Ansible will not keep the network connection alive and will give the Grace Blackwell chip the full 60+ seconds it needs to process the firmware layers, serialize its memory planes, and finish calibrating and setting up the high-speed links. Once the links are bound, Ansible safely re-authenticates.

P.S. (Written by the human author, not the AI): Strictly speaking, you do not need the final ansible.builtin.wait_for_connection: step (it even slows down your script), but it gives you the confirmation that the Spark DGX gained network connectivity after the reboot.

The Mitsubishi Outlander PHEV is a nice and very reliable EV, but it has one issue: battery degradation is not measured but estimated, assuming loss of battery capacity based on the calendar, but also based on charging cycles. Typically, this estimate is too pessimistic and therefore, the Battery Management Unit (BMU) does not allow to use fully the real battery capacity.

However, using a so called Drive Battery Capacity Automatic Measurement (DBCAM), the BMU will drain the battery to 0% (by switching on air condition and heating simultaneously which is a good way to consume significantly energy without actually driving) and after that, the battery gets fully charged and while charging, the BMU measures how many energy is charged. This measured value replaces then the estimated battery capacity, giving you a some range back.

The following describes using the Thinkcar Thinkscan 689 diagnostic tablet to do this. This diagnostic tablet has the advantage that it can be used without an expensive subscription. While the diagnostic tablet needs to be bought once, it pays of in comparison to paying your car dealer every year to the DBCAM.

Also the app PHEV Watchdog lite is used (together with an OBD Bluetooth adapter) to read and store BMU data.

Cell smoothing

It might that doing a cell smoothing before the DBCAM gives even better DBCAM results! But whether this is worth it depends on the min/max voltage of the cells: only if the delta is 0.050V or more, its worth it. For assessing the cell drift, use PHEV Watchdog lite and it is best to do this with an empty battery.

However, for a cell smoothing, the batteries need to be full! (Just like a normal cell balance at the end of charging. But the difference is that cell balancing charges the cells that are not yet full, using the tiny resistors for those cells that are already full, i.e. I charges cells so that they reach the voltage of the fuller cells. Whereas the smoothing uses the tiny resistors to discharge those cells that have a higher voltage than the other cells.) So, first do fully charge before you start a cell smoothing.

Connect: Plug your Thinkcar diagnostic tablet into the OBD port while the vehicle is fully OFF.

Ignition to READY: Turn the vehicle on to READY mode (foot on the brake, press power). Note: This is different from the DBCAM, which requires the car to just be in the "ON" accessory mode.

Navigate: Open the Thinkcar app, decode your VIN, and navigate to System Selection -> BMU.

Execute: Select Special Function -> Cell voltage smoothing.

Confirm: Press the button to execute the command.

Monitor: The Thinkcar screen will show an "Execution status." You simply wait until that status changes from "Active" to "Not Active" to confirm the balancing is completely finished.

Time: The process can take up to a maximum of 2 hours to complete.

Tablet Power: Your Thinkcar tablet must not go to sleep or turn off its screen during this process. If it does, the connection will drop, and the procedure will abort and throw an error. Plug your tablet into a charger and set its display settings to stay awake indefinitely.

DBCAM

You may find on the web a Mitsubishi Service Bulletin MSB-17EXL54-501 (filename: MSB-17EXL54-501.pdf) that describes the procedure using the Mitsubishi M.U.T.-III diagnostic tablet. The following is based on that description:

If you like, connect PHEV Watchdog lite in order to make there a new reading of the current battery condition (connect via Bluetooth, then via three dots: Battery Condition -- when connected via Bluetooth, it should immediately record a new reading. By tapping on it, you can delete a reading in case you made too many recordings of the same state).

1. Drive the drive battery to empty
2. Have car switched off
3. Connect charging cable: car will start charging (even though we want to have the battery empty, but as it is anyway a slow charger, it is no problem that this will now actually start charging the car)
4. Connect the diagnostic tablet (I had the car switched on, but I had later, in step 17., to switch my car off for the DBCAM to start: try next time to leave it switched off! The Mitsubishi Service Bulletin MSB-17EXL54-501 also mentions that the car should be switched off before connecting their diagnostic tablet.)
5. Do the VIN scan: click the large VIN button
6. Click: Automatically Search. It will then show Vehicle information, click: OK
7. Click: System Selection
8. Scroll down to and click: BMU
9. Scroll down to and click: Special Function
10. Scroll down to and click: Battery Auto Capacity Measured, confirm with: Yes
11. As DBCAM is currently not running, but maybe completed some earlier run, Execution Status should be either: Not run (if you never did a DBCAM) or Complete (from the previous DBCAM)
12. Scroll down: Auto Capa.Measured Result HIS.1 to 10 should show the history of previous DBCAMs battery capacity measurements (or the original capacity, 40 in case of my MY 2017), Auto Capa.Measured Result HIS. 1 to 10 should show the odometer readings from previous DBCAMS, or 0).
13. Note down what value is shown for Battery Current Capacity -- after DBCAM completed, it should have changed (and will hopefully be a higher number as it is using the measured instead of the too pessimistically value calculated by the BMU).
14. At the very bottom of the list, Capa. Measured Discharge Current should show NG/Not Performed
15. At the bottom right, click: Perform (to start the DBCAM) and confirm with: yes. It should then say: Successful Operation, press: OK
16. Execution Status should now be: Executing
17. As I had my car still switched on, at the very bottom of the list, Capa. Measured Discharge Current was still showing: NG/Not Performed
18. Finally, after switching off my car, the fan, AC, heater, turned on (i.e. everything that consumes energy in order to empty the drive battery to 0%). The diagnostic tablet was still running and did then display in the list at the very bottom, Capa. Measured Discharge Current was still showing: Ok
19. As the car was in my garage, I did not lock it, but it should be possible to lock the car now.

As the fan and heater are running to drain the battery, this means: Success. Emptying the drive battery to 0% will now take some hours (aprox. 1-2 h). All documentation says that is is now possible to disconnect the diagnostic table (and connect later again to check).

I did let the diagnostic table connected, but after some time, the car put anyway the OBD port to sleep, so the diagnostic tablet did disconnect. I left my car in my garage, i.e. I did not lock the car doors.

Sooner or later, the battery is empty and fan, AC, etc. should go off. The car should now have automatically started a new charging session, i.e. charge the empty battery (and while doing that, it will measure how much it is able to charge the battery to determine the real capacity): the red charging symbol in the dashboard should be on.

While it was charging, I did re-connect my diagnostic tablet to the car:

1. I did the above VIN scan, but when selecting: Automatically Search, it did return an error and I had to manually enter the VIN (even though it did just display the VIN a few seconds before).
2. After that, I was able to select: System selection
3. BMU
4. Special Function
5. Battery Auto Capacity Measured
6. Execution Status should still show: Executing
7. Capa. Measured Discharge Current should show again: NG/Not Performed

When this measured charging session finished (will take longer than a normal charge that would start at ca. 30%, as this time the charge starts at 0%), the red charge light in the dashboard should go off.

I then connected the diagnostic table again and switched on the car.
I still had the charging cable connected and interestingly the 12 V battery light was on, i.e. the OBC was not charging the 12 V battery, but the car was in accessory mode (I am pretty sure that I had pressed the brake pedal while switching the car on). I then switched the car off, disconnected the charging cable and then, switched the car on: the 12 V light was gone.

1. I did the usual VIN scan, and Automatically Search did work again this time.
2. After that, I was able to select: System selection
3. BMU
4. Special Function
5. Battery Auto Capacity Measured
6. Execution Status should be: Complete,
7. Error status should still be: No error
8. Capa. Measured Discharge Current should show again: NG/Not Performed
9. Battery Current Capacity should show a new value (hopefully higher than the old value).
10. Also the HIS 1 value should have been shifted to HIS.2 and so on.

The DBCAM is now finished and you can disconnect the diagnostic tablet. If you like, connect PHEV Watchdog lite. It should say that a new capacity measurement is available and via Battery Condition, the DBCAM should be shown and the new (hopefully better) capacity should be shown.

TPMS Valves programming

Thinkcar also has a device for reading TPMS valves, transfer that via Blueetoth to the diagnostics tablet and then writing the valve ID via the diagnostics tablet to the Outlander. That TPMS reader might only work with Thinkcar valves, though?

Update: All notifications have been sent. The tutorial and the workshop have been cancelled, workshop papers have been integrated into the main track.

Update (12th of May 2026): For some of the papers submitted to the second submission round, we have today sent out notifications. But others are still under review, so please be patient if you submitted to the second submission round and did not yet get a notification. While some accepted papers are already listed (e.g. from the first submission round), if you did not yet get a notification, then your paper is still under review.

Update (7th of May 2026): we are late with notifications for the second submission round that was planned to be 7th of May 2026.

A late track submissions has been opened with new deadlines (this is strictly speaking not a deadline extension as the submissions of the first two rounds will be reviewed already now and notificated as planned):

Late track Submission deadline: 25th April 2026

Late track Poster submission deadline: 25th April 2026

Late track Notification: 17th May 2026

Late track Camera ready: 30th May 2026

Also notice the Workshop Trust and Deception as Security Primitives that has still submission open.

In addition, there will be a tutorial: Autonomous Drone Systems Cybersecurity and Threat Intelligence

The 17th EAI International Conference on Digital Forensics & Cyber Crime (ICDF2C 2026) will take place in Reykjavik, Iceland, 8 - 10 September, 2026. Hosted by the Computer Science department of University of Iceland. Proceedings published by Springer.

For more details, check the original call.

CALL FOR PAPERS Late track extended deadline (Apologies for cross-posting)

EAI ICDF2C 2026: 
17th EAI International Conference on Digital Forensics & Cyber Crime
publication by Springer

https://icdf2c.eai-conferences.org/2026/

When: 8 - 10 September, 2026

Where: Reykjavik, Iceland

Late track Submission deadline: 25th April 2026

Late track Workshop submission deadline: 25th April 2026

Late track Notification: 17th May 2026

Late track Camera ready: 30th May 2026


Scope

The 17th EAI International Conference on Digital Forensics & Cyber
Crime (ICDF2C) will be held on 8-10 September, 2026, in Reykjavik
(Iceland). This three-day event is expected to attract over 100
participants, including academics, practitioners, criminologists (or
law enforcement) and vendors, providing business and intellectual
engagement opportunities among attendees. The conference is organized
by the European Alliance for Innovation (EAI).

This conference's theme is cyber analytics and forensics in the era of
emerging threats. Novel cyber threats are continuously emerging,
catalysed by the rapid deployment of Large Language Models and other
AI across many domains which increases the threat surface in many
sectors such as Smart Industry, Fintech and digital government. The
focus of this conference is to provide a platform for discussing these
emerging threats and to identify priorities for the community to
target with the next generation of cyber analytics. We particularly
welcome research which studies the dynamics between human factors and
AI technologies and the corresponding impact upon cybersecurity and
forensics.

We encourage the authors to use the Posters and Demos venue as a way to 
open up discussions with the ICDF2C community about their early work in 
progress and develop the work for future collaborations. 
Representatives from industry, including established companies and 
startups, are warmly welcome to showcase products and services that 
are related to the topics of the conference.


Workshop: Trust and Deception as Security Primitives
Social engineering remains one of the most effective attack vectors 
against modern socio-technical systems, yet it is still poorly integrated 
into formal security engineering practice. This workshop advances a 
unifying perspective that treats trust and deception as first-class 
security primitives—measurable, attackable, and engineerable system 
properties across human, organizational, and AI-mediated environments.

The workshop brings together researchers and practitioners working on 
adversarial modeling, trust measurement, AI-mediated influence, 
detection systems, and human-in-the-loop defenses. 
Participants will examine how social engineering threats can be 
rigorously modeled, empirically evaluated, and mitigated using 
computational, organizational, and policy-aware approaches.

Through paper presentations and structured breakout discussions, 
the workshop aims to (1) consolidate fragmented research threads 
into a coherent socio-technical security framework, 
(2) identify shared metrics and evaluation challenges, and 
(3) define a research roadmap for defensible, trustworthy systems 
in high-consequence domains. 


Tutorial: Autonomous Drone Systems Cybersecurity and Threat Intelligence
An intensive 4-hour program exploring critical intersections of 
autonomous drone technologies, cybersecurity vulnerabilities, and 
strategic threat intelligence. Designed for cybersecurity professionals, 
intelligence analysts, and technology researchers seeking advanced 
insights into drone system exploitation and defensive strategies.

***

Publication

All registered papers will be submitted for publishing by Springer –
LNICST series and made available through SpringerLink Digital Library:
ICDF2C proceedings.

Proceedings will be submitted for inclusion in leading indexing
services, such as Web of Science, Compendex, Scopus, DBLP, EU Digital
Library, IO-Port, MatchSciNet, Inspec and Zentralblatt MATH.

All accepted authors are eligible to submit an extended version 
in a fast track of:
- EAI Endorsed Transactions on Security and Safety
- EAI Endorsed Transactions on Internet of Things

***

Topics

Theme: Cyber analytics and forensics in the era of emerging threats.

Applications of artificial intelligence (AI) and other related technologies:
- Anti-forensics and anti-anti-forensics (e.g., deepfake)
- Deep learning
- Explainable AI (XAI)
- Generative AI (GenAI)
- Large language model (LLM)

Device forensics:
- Blockchain investigations
- Internet of Things (IoT) forensics 
 (including industrial IoT, medical IoT, military IoT, battlefield IoT, and vehicular IoT)
- Edge and/or cloud forensics
- Network and distributed system forensics
- Virtual / augmented reality (VR/AR) forensics
- Other emerging / contemporary technologies 
 (e.g., hardware and software such as firmware and operating systems)

Financial crime investigations:
- Financial frauds and scams
- Cryptocurrency investigations
- Market manipulation investigations
- Anti-money laundering / counter terrorism financing investigations
- Anti-corruption investigations

Cyber security and analytics:
- Network security (e.g., intrusion detection)
- Malware analysis
- IoT security
- Security operations center
- Virtual / augmented reality (VR/AR)

Education and Evaluation:
- Case studies – legal (e.g., child sexual abuse material) and/or technical
- Infrastructure
- Methodology
- Replicability and validity
- Tool validation

Theory and fundamentals:
- Anti-forensics and anti-anti-forensics (e.g., encryption and deepfake)
- Frameworks (legal, policy, and/or technical)
- Privacy-preserving forensics
- Social and privacy
- Steganography and steganalysis
- Visualization methods and tools for forensic analysis

***

General Chairs
Helmut Neukirchen - University of Iceland, Iceland
Kim-Kwang Raymond Choo - University of Texas at San Antonio, USA

Technical Program Committee Chairs
Thomas Welsh - University of Iceland, Iceland
Hans P. Reiser - Reykjavík University, Iceland
Raymond Chan - Singapore Institute of Technology, Singapore

***

This event is organized by EAI https://eai.eu/

EAI – European Alliance for Innovation is a non-profit organization
and a professional community established in cooperation with the
European Commission to empower the global research and innovation, and
to promote cooperation between European and International ICT
communities. 

The deadlines for applying for our Computer Science, Software Engineering and Computational Engineering programmes are as follows:

• M.Sc.: Deadline 15. April
• B.Sc.: Deadline 5. June
• International students: deadline was 1. February -- except for students from the Nordics: here the same deadline as above apply

As I was involved in updating our programmes, I can assure you that you will receive up-to-date education and can really recommend studying with us.

It was in the news that Iceland will need 9000 specialists within next 5 years -- you will be one of them if you studied at University of Iceland. While due to the advent of AI, it is often said that only senior programmers are needed, but not junior programmers anymore -- but where should senior programmers come from if they were not junior programmers before? Also, someone needs to develop the AI models: this is exactly what you learn at University of Iceland where you can take course on AI, including a specialisation in Artifical Intelligence/Machine Learning (Gervigreind).

Our B.Sc. study programmes are: B.Sc. in Computer Science/Tölvunarfræði and B.Sc. in Software Engineering/Hugbúnaðarverkfræði.
In Computer Science, we have additional specialisations in AI/Machine Learning (Gervigreind) and Computational Science (Reiknifræði).

Our M.Sc. study programmes are: M.Sc. in Computer Science/Tölvunarfræði, M.Sc. in Software Engineering/Hugbúnaðarverkfræði, and M.Sc. in Computational Engineering.
We have additional specialisations in Cybersecurity (Netöryggi), Language Technology (Máltækni) and Innovation (Nýsköpun og viðskiptaþróun -- the Innovation courses are taught in Icelandic).

If you want to learn more about Software Engineering, I have a blog post Why you should study Software Engineering / Af hverju hugbúnaðarverkfræði and a distinction of Software Engineering versus Programming.

Update: A second round for submissions has been opened with new deadlines (this is strictly speaking not a deadline extension as the submissions of the first round will be reviewed already now and notificated as planned):

2nd round Submission deadline: March 27 2026

2nd round Notification: April 30 2026

2nd round Camera ready: May 15 2026

The 17th EAI International Conference on Digital Forensics & Cyber Crime (ICDF2C 2026) will take place in Reykjavik, Iceland, 8 - 10 September, 2026. Hosted by the Computer Science department of University of Iceland. Proceedings published by Springer.

For more details, check the original call.

University of Iceland will be present at UTmessan 2026, both on the experts conference day on Friday, 6. February 2026, and on the public day, Saturday, 7. February 2026. On the conference day, research will be presented -- also from Reykjavik University, as part of our cybersecurity collaboration, e.g. the Frostbyte lab. The public day will be a mixture of student work being presented and showcasing Gagnabær ("Datatown"), a digital twin that visualises cyber attacks on our server in Iceland using a LEGO model.

---

This event is in the context of our cybersecurity activities and the ECCC/EU co-funded projects ICEDEF – Defend Iceland and Eyvör – the National Cybersecurity Coordination Centre of Iceland (NCC-IS).

Join us for an afternoon of exploring different aspects of Internet security! National and international experts from various sectors will gather to share their insights from a wide range of Internet security-related topics. We envisage an afternoon full of information-sharing, lively discussions and networking opportunities for the Icelandic Internet community.

Organised by ICANN, hosted by ISNIC, advertised by University of Iceland.

Agenda
14.00 – 14.05 Welcome from ISNIC & ICANN, Þór Jensen, ISNIC and Ulrich Wisser/Gabriella Schittek, ICANN

14.05 -14.25 Migrating ISNIC’s DNSSEC Keys to a New Algorithm, Einar Bjarni Halldórsson, ISNIC

14.25 – 14.50 ICANN's Work on DNS Abuse - Ulrich Wisser, ICANN

14.50 – 15.15 How the Icelandic Police Combats Cyber Crimes, Steinarr Kr. Omarsson, Reykjavik Metropolitan Police

15.15 – 15.45 Break

15.45 – 16.10 Coordinating the response to an ever-evolving phishing group, Hafsteinn Baldvinsson, CERT-IS

16.10 – 16.30 After the Click: The Race to Contain and Respond to Threats, Bergsteinn Karlsson, Ambaga

16.30 - 16.50 Surveillance and Security Requirements in the Era of NIS2 - Unnur Kristín Sveinbjarnardóttir, Cybersecurity Division, Fjarskiptastofa / Electronic Communication Office of Iceland

16.50 – 17.10 Threat Led Penetration Testing, Kristján Valur, Jónsson, Seðlabanki Íslands/Central Bank of Iceland

17.10 - 17.30 Lost While Protected: Privacy As an Alibi, Lára Herborg Ólafsdóttir, LEX

17.30 - Social event Beer Garden, Fosshotel Reykjavik

Venue
ISNIC
Katrínartún 2
Reykjavik

Registration needed