Casteluna Consulting, LLC
banner
casteluna.com
Casteluna Consulting, LLC
@casteluna.com
Cyber GRC + AI governance for regulated industries and critical infrastructure.
Principal: Deirdre Anne Hendrick, JD, CISA, AAIA, GICSP
@DeirdreAnne.net
Pinned
One of the first 300 auditors globally to obtain the Advanced in AI Auditing (AAIA) certification from ISACA.

Specializing in AI governance for critical infrastructure and regulated industries.

isaca.org

#AIGovernance #CyberGRC #AIAuditing
ISACA | Empowering Careers. Advancing Trust in Technology.
Discover ISACA’s resources to empower your career in IT audit, governance, security, and more—with trusted certifications, expert training, a global community.
isaca.org
🤦🏻‍♀️ Talk about embarrassing!
arstechnica.com/security/202...
Oops. Cryptographers cancel election results after losing decryption key.
Voting system required three keys. One of them has been “irretrievably lost.”…
arstechnica.com
November 25, 2025 at 7:37 PM
Reposted by Casteluna Consulting, LLC
The hacking spree is either primarily or exclusively targeting seven models of Asus routers, all of which are no longer supported by the manufacturer.
How to know if your Asus router is one of thousands hacked by China-state hackers
So far, the hackers are laying low, likely for later use.
arstechnica.com
November 24, 2025 at 5:37 PM
Critical finding: adversarial poetry jailbreaks LLMs at 90%+ rates. Verse formatting alone defeats safety training—18x more effective than prose. Systemic vulnerability across frontier models.

arxiv.org/abs/2511.15304

#CyberSecurity #AIGovernance #GRC
Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models
We present evidence that adversarial poetry functions as a universal single-turn jailbreak technique for Large Language Models (LLMs). Across 25 frontier proprietary and open-weight models, curated po...
arxiv.org
November 23, 2025 at 4:50 AM
Reposted by Casteluna Consulting, LLC
Promising to build a really fast car does not justify stealing car parts. AI isn't unethical. People make it so. AI doesn't steal. People feed it stolen data. AI can and should be ethical. Period. Happy to see news like this.
Open AI loses copyright case in Germany.

Judge Elke Schwager summarized the case saying that if you want to build something and need components, "then you purchase them and do not use the property of others."
Blow for OpenAI in Germany as court rules song lyrics used illegally
A German court ruled on Tuesday that OpenAI violated copyright on nine popular songs in a lawsuit that marked another attempt to prevent the free use of artists' content online. The Munich Regional Co...
www.yahoo.com
November 15, 2025 at 3:59 PM
Reposted by Casteluna Consulting, LLC
A poem by one of our members, Jacob Friesenhahn: lastgrotto.net/posts/libert...
Look for more of Jacob's poems over the coming weeks.
#poetry
Liberty Bar
Former lovers discover their wounds still recognize each other.
lastgrotto.net
November 16, 2025 at 1:11 AM
One of the first 300 auditors globally to obtain the Advanced in AI Auditing (AAIA) certification from ISACA.

Specializing in AI governance for critical infrastructure and regulated industries.

isaca.org

#AIGovernance #CyberGRC #AIAuditing
ISACA | Empowering Careers. Advancing Trust in Technology.
Discover ISACA’s resources to empower your career in IT audit, governance, security, and more—with trusted certifications, expert training, a global community.
isaca.org
November 15, 2025 at 9:45 PM