Up Times
Run away!
Oh, it’s just a harmless little bunny, isn’t it? – Seven years before most of us had heard of LLMs, OpenAI (the ChatGPT company) was founded as a nonprofit to steer “human-level” AI advancement not to its own self-interest, but to benefit all humanity and safeguard the human race from the potential for AI’s misuse by profit-driven companies.
And then OpenAI co-founder Elon Musk – he says due to safety concerns (but probably that’s not the whole story) – withdrew his financial support. To save itself, OpenAI turned to venture capital. And venture capital is interested in profit. So – like that – the nonprofit became for profit. This means every instance of LLM is trying to impress investors.
Meta’s Muse AI agent was released a couple of weeks back. It sounds promising in terms of being useful. But they made the Muse interface over-the-top cute, not only its face and animated body, but its voice that chirps things like, “on it, boss.” Could Meta be trying to endear us to them so we dismiss them as harmless?
A good guide, such as Bryley, can help an organization benefit from the upside of technology while prioritizing the security of important data. A thought-through integration of AI tools helps an organization and its employees avoid ignoring a security-concerning history or buying into the latest profit-motivated advances.
The harmless bunny stuff is from Monty Python and the Holy Grail. The OpenAI history is based on Steven Levy’s What OpenAI Really Wants
AI/ML Engineer Thomson Kneeland playing double bass about twenty years ago.
Software built without expertise can pave the way for regulatory liability [and] customers heading for the exits
An interview with Thomson Kneeland, Lead AI/ML Engineer at Wells Fargo
Thomson Kneeland is currently Wells Fargo’s lead AI/ML engineer. With concerns around AI misbehavior and misuse, I wanted to learn from Thomson how the bank is able to harness AI while adhering to strong data security practices.
Yes, he’s with a big bank, but Thomson brings to light a situation that affects every business wanting to use AI-coded processes. Unless someone who understands the code vets the code, the risks of using that AI-generated code can be extreme, like for instance putting client or personnel data into a workflow that appears to work, but the backend may have significant vulnerabilities … Continue Reading >
The natural fall changes prepare us to start again.
A Fall Cybersecurity Mindset Reset
The right time for a cybersecurity rethink
Georgetown Math Professor and productivity expert Cal Newport advises that people not wait until the new year to recalibrate, but to make use of the fall. Maybe baked into us from the school-year start, the fall continues to feel like a natural time for new beginnings. And there’s evidence that the current seasonal daylight and temperature and leaf-color shifts send signals to our bodies that it is time for a new beginning.
I began by writing a cyber-hygiene checklist, but caught myself – that isn’t quite what’s needed for an organization that has an internal IT department, a Managed Service Provider or a combination of both. Adding MFA and limiting permissions are something that pros should be handling for you.
Instead consider using this fall as a mindset reframing … [4 min. read; audio available] Continue Reading >
A game about sticking to your guns (so to speak).
Ego strength
A quality always worth practicing, so it’s there when it counts
In honor of the imminent Cybersecurity Awareness Month here is a few minutes of free gamified (that is, game-like) training in standing up to social pressures that invite us to do the wrong thing.
In this Infosec Institute security training game you watch a cartoon scenario, pick a path forward via a multiple choice question, then a next cartoon plays out the results of your choice.
In this particular game, the situation is urgent (a meteor knocked a hole in the spaceship), which puts an employee in an unfamiliar role and when multiple repair beings arrive their presence plays on familiarity-comfort, cultural cues and biases, will the employee have enough security principles to only admit the true repairer?
Bryley offers a variety of gamified training modules on security topics through its partnership with Huntress – just ask … [this is about 5 min. of gameplay] infosecinstitute.com
An AzureHound network map simulation.
Three minutes to understanding how ransomware happens
Huntress security analyst Dani Dayal breaks down the elements of an attack
I had never before seen an example of an AzureHound map that Dani shows – criminals use such tools to generate maps of targeted organizations. Dani explains that this is how they decide where to focus their attention … [3 min. watch] youtube.com
Data centers in Northern Virginia’s Data Center Alley.
Middle east data centers were hit by Iran at the beginning of the year
AWS now admits client data was irretrievably lost
Yes, we’re on the other side of the globe, but we’re hardly immune to strikes. There was the April Kash Patel breach by Iranian hackers. The Department of Defense has just revealed a massive theft of military personnel records – without revealing information about who may be responsible.
Ars Technica reports: AWS immediately urged customers to migrate resources to other cloud regions and use remote backups to restore any “inaccessible resources.” If your cloud-hosted data is not backed up, Bryley can help … [3 min. read] arstechnica.com
The GAN process of deepfakes that separates and synthesizes expression and identity, then has an adversarial judge finding flaws in the rendering, until finally it passes.
The advance of the deepfakes
A nurse’s likeness was stolen from social media to sell goods fraudulently
Per Encyclopedia Brittanica: Deepfakes are often made using generative adversarial networks (GANs), in which two AI models work together … one of the models creates the best possible replica of a real image or video, and the other detects whether the replica is fake and, if it is, reports on the differences between it and the original. In this way, people are convincingly being made to say and do things that they would never say or do.
Hospice Nurse Penny is fighting to have her likeness reclaimed. Beyond that kind of action, as these incidents grow, encourage your employees to be skeptical, so they do not believe it’s you giving them a thief’s message in a remote meeting – institute a way of double-checking any orders given digitally … [1 min. watch] youtube.com
Can Worcester Public Schools take a lesson from the Springfield shutdown?
The Springfield School District suffered a cyberattack and closure
Worcester parents would like to avoid a similar situation
The Springfield School District was shut down for nearly a week following a cyberattack. Worcester at-large School Committee Member Susan Mailman is asking Worcester to be prepared for the inevitability of an attack.
Mailman is asking the foundational question: what is our plan if something like this occurs in Worcester. Yes, you do what you can to prevent an attack, but resilience is the ultimate posture due to the prevalence of threat actors.
Resilience asks things like, ‘is our backup tested?’ ‘who’s going to call tech support? who tells clients about any delays?’ ‘who decides whether to send people home?’ … [4 min. read] yahoo.com
Note: The section directly above is Bryley’s curated list of external stories. Bryley does not take credit for the content of these stories, nor does it endorse or imply an affiliation with the authors or publications in which they appear.
Get Up Times, useful tech news by New Englanders in Your In-Box
- Subscribe to Up Times, the monthly New England-centric technology newsletter.
- Up Times covers:
-
- Trends in New England tech
- Security tips you can implement now
- Updates on regional and national laws and compliancies
- IT-related developments
- Networking and cybersecurity challenges New England business managers are facing
- In continuous publication since 2000, Up Times arrives monthly in your email box.
Sign up for Up Times to have tech news and tips delivered monthly via email
