iiinova — where your digital self keeps evolving

Kami Vaniea @kaniea.bsky.social · 28/10/2025

Sirisha Rambhatla presenting at the @uwaterloo.ca cybersecurity and privacy institute industry day. Building efficient and safe LLMs.

110

Kami Vaniea @kaniea.bsky.social · 20/10/2025

Incidents like this make patch management challenging. Windows patch broke mouse and keyboard support in *recovery mode*. Normal operation still fine, but if anything breaks.... recovery is now gone. How many admins even test for this? www.tomshardware.com/software/win...

tomshardware.com

Windows 11's October update just broke the Windows Recovery Environment — USB keyboards and mice unusable in Windows RE after latest bug hits

Another week, another OS-breaking bug.

Kami Vaniea @kaniea.bsky.social · 09/09/2025

Canadian Anti-Scam Coalition set to launch this afternoon, its an alliance between credit card companies, banks, telecoms, Canadian government, and tech companies. The aim is to share information and reduce fraud. thelogic.co/news/big-ban...

thelogic.co

Canada’s big banks, telecoms and tech firms are joining forces to fight scams - The Logic

From Bell to Google, major firms in Canada are following Australia’s model of sharing information to fight fraudsters

120

@lujobauer.bsky.social starting off USENIX security to a large packed room.

020

Closing #soups2025 remarks by our incredible general chair Patrick Gage Kelly.

130

Clement Fung presenting about industrial control systems (ICS). They interviewed people who protect ICS. Current alarm systems have a great variety in how they are designed and even who is in charge of managing them. #soups2025 www.usenix.org/conference/s...

100

Nina Gerber @ninag72.bsky.social presenting on social and emotional dimensions of security. How can security and privacy be made more enjoyable? Used sentence completion tasks. People found important but not interesting. #soups2025 www.usenix.org/conference/s...

100

Qiurong Song presenting on predatory monetization in video games such as Roblox. Loot boxes being a example. Developers are adding predatory design on purpose. Use of possess and in-have currency make it harder to stay aware of how much is spent. #soups2025 www.usenix.org/conference/s...

000

Teens and privacy in terms of barrier management. Navigate self disclosure. Issues like ambiguous norms. Also the ability to trust an audience. #soups2025 www.usenix.org/conference/s...

011

Jenny Tang presenting about doing and reporting statistics in research papers. #soup2025 There are many ways to do statistics incorrectly and even if the test is correct the necessary numbers to interpret them may not be provided. @lujobauer.bsky.social www.usenix.org/conference/s...

111

Annalina Buckmann (soon on job market) presenting on sociodemographic factors and barriers to security and privacy adoption. Mapped findings to SPAF factors. #soups2025 www.usenix.org/conference/s...

001

Anna-Marie Ortloff presenting a replication of the "No one can hack my mind" paper. #soups2025 Experts have even more focus on authentication. End users also using 2FA increasing anf anti virus usage going down. www.usenix.org/conference/s...

001

Laura Marie Abels presenting on Social Desirability and developing a scale to measure it. #soups2025 www.usenix.org/conference/s...

001

Lily Klucinec presenting about mobile money in Kenya which is a financial system tied to SIM cards. To withdraw money personal information currently have to be shared with a mobile money in-person agent. #soups2025 @lorriecranor.bsky.social www.usenix.org/conference/s...

031

Time based one time passwords, such as those generated by Google Authenticator app. Part of the FIDO Alliance. The paper compares physical secury keys to TOTPs+passwords. #soups2025 www.usenix.org/conference/s...

101

Peter Mayer presenting about forest time users of password managers #soups2025 Before the study on average had 36 passwords, some of which were reused. There was use of the manager such as adding new passwords. www.usenix.org/conference/s...

101

Analysis of NIST authentication advice in standards documents. #soups2025 Multi factor advice is followed by many universities. Password regular password changes (no longer recommended) still common. www.usenix.org/conference/s...

101

People take videos of strangers and then use TikTok to identify the person. Many people are identified this way. While sometimes heartwarming, it can also be used for abuse and stalking people. #soups2025 www.usenix.org/conference/s...

021

Usability of privacy nutrition labels. #soups2025 Terms in privacy labels are widely misunderstood. www.usenix.org/conference/s...

011

Mark Wong presenting about essential online services and minoritised ethnic people. #soups2025 Concerns include control of personal data. People were being selective in what they share. They also make up days. That also don't use some services out of worries. www.usenix.org/conference/s...

111

Amna Batool presenting on between court orders and platform policies at #soups2025 Non consentual image disclosure, even seemingly innocent images can in some cultures lead to violence. www.usenix.org/conference/s...

100

#soups2025 day 2 about to start. www.usenix.org/conference/s...

000

#soups2026 keynote Greg Chappell on "Mind the gap between research and practice"

Playing in the Sandbox: A Study on the Usability of Seccomp | USENIX

Joseph Hallett presenting on behalf of Maysara Alhindi at #soups2025 How do developers go about adding sandbox to code. All participants ran strace as a start point. Using chatgpt made it take longer. Everyone took a different approach. www.usenix.org/conference/s...

usenix.org

100

Yara Alsiyat presenting on privacy, IoT, and cultural sensitivity at #soups2025 Smart cameras are seen as a social presence, capable of violating privacy. It would be helpful to have a guest mode, and pausing for times like prayer. www.usenix.org/conference/s...

000

Kieron Ivy Turk presenting on IoT abuse at #soups2025 Human harms model developed to categorise harms. This study looks at how people learn how to use IoT devices in abuse situation. www.usenix.org/conference/s...

110

CheckEngine, testing embeded browsers in smart TVs. Asked people to visit a special link via their smart TVs. Generally security policies were implemented correctly. But they were not updated. #soups2025 www.usenix.org/conference/s...

100

AilIsafety4kids.org youth-genAI tools presented by Yaman Yu at #soups2024 Youth are using GenAI in ways parents might not expect. Research data is available. www.usenix.org/conference/s...

100

LLMs in security incident response. #soups2025 An incident report is the final stage of an incident response. LLMs are good at summarising documents. Can they do a good job in this case? www.usenix.org/conference/s...

100

Stephen Meisenbacher presenting on AI developers and privacy risks. #soups2025 Interviewed 25 AI developers. Amplification factors were often mentioned - increased accessibility, capabilities, data usage. www.usenix.org/conference/s...

100

Writing good survey questions that accurately assess safety, trust, risk, and fairness is challenging. #soups2025 www.usenix.org/conference/s...

100

"You go now! No trouble!" Understanding the Offboarding Process in Companies from an IT Security Perspective | USENIX

IT Offboarding: what happens when employees leave.... #soups2025 Ideally when an employee leaves a company outsiders like access to systems should be correctly handled. Offboarding is handled by people with a wide range of job titles. www.usenix.org/conference/s...

usenix.org

100

Scan discussions on Reddit #soups2025 Preventative communities that try and help people and engagement communities which try and waste scammer time. They find Reddit to be supportive of victims. Framing situation as a professional attack by experienced attackers. www.usenix.org/conference/s...

100

Phishing via SMS messages can be hard to detect and the current earnings do not help users build good mental models. A new LLM is built to help classify messages and provide explanation to users: SmishX www.usenix.org/conference/s...

100

Language as Lure: A Naturalistic Study on Pasifika Phishing Susceptibility | USENIX

Phishing up next: Language as Lure: A Naturalistic Study on Pasifika Phishing Susceptibility. Linguistic and cultural aspects can impact people's likelihood to respond to phishing. #soups2025 www.usenix.org/conference/s...

usenix.org

100

Kami Vaniea @kaniea.bsky.social · 10/06/2025

Starting off #soups2025 paper presentations: Rajvardhan Oak on Pig-Butchering Scams.

201

"The Cybersecurity Bonfire of Human Capital" <- great title by M. Angela Sasse at SHB workshop.

110

Kami Vaniea @kaniea.bsky.social · 15/05/2025

I briefly turned on Recall to play with it a bit. A few observations: * If a filtered window is on top, it screenshots the invisible-to-user window under it * Does not screenshot Firefox at all, and therefore screenshots anything under Firefox * Turn on/off links vanish from Recall UI after turn on

000

Kami Vaniea @kaniea.bsky.social · 15/05/2025

Windows machine asked to turn on Recall this morning. I said no, but still got asked a second time and an icon auto added to the taskbar in case I want to change my mind. They must really want people to turn it on.