Lightnews — Scholar-powered news

Bryley @bryleysys.bsky.social · 3d

October's Up Times newsletter to read and listen to bryl.link/s83 – including: why this month is a good time for cybersecurity awareness + from MIT: the problem of humanoid-form robots; also navigating the data issues of Microsoft/OpenAI's changing relationship #netechnews #cybersecurityawareness

Bryley @bryleysys.bsky.social · 12d

1000s of ChatGPT chats showed up in Google Search, including w personal & work info. Even if a chat tool feels private, it's often not: "the internet doesn’t forget, even when you try to delete," per Tom's Guide bryl.link/ia7 #privacy #chatbots

1

Bryley @bryleysys.bsky.social · 18d

Guide to the Uptime Mindset: 'fix it when it breaks' works for pencils and pens, but the stakes are different when operations are built on tech tools. Learn more: bryl.link/wq8 #Uptime #uptimemindset

Bryley @bryleysys.bsky.social · Sep 18

Did you know cyberattacks can be detected before they cause damage? SOCs are command centers where IT experts monitor networks 24/7, analyzing threats in real-time. SOC teams can integrate with Endpoint Detection and Response ... [7 min. read] bryl.link/hpz #soc #edr

Bryley @bryleysys.bsky.social · Sep 16

September Up Times newsletter: cultivating the uptime mindset, Jen Easterly on tech-issues responsibility, famous parallel CPUs and a famous designer using parallel GPUs bryl.link/utn #itservices #newenglandit

Bryley @bryleysys.bsky.social · Sep 16

Why do we cut security corners despite the risks? Concentrated benefits (password reuse, ignoring best practices to save $) vs diffuse costs (distant consequences). Those who benefit directly from shortcuts outweigh broader security concerns bryl.link/7pb #cybersecurity

1

Bryley @bryleysys.bsky.social · Sep 15

OpenAI's GPT-5 still vulnerable to prompt injection attacks (56.8% success). Hidden malicious instructions in documents can manipulate AI responses, ex: telling chatbots to recommend removing liability protections as standard practice ... bryl.link/4vb #AI #Security

Bryley @bryleysys.bsky.social · Sep 11

We'd seen many bad guys each trying to break in a house. Now it's looking like an organized cyber-syndicate to watch a target's routines and share info with the intention of keeping at it till the mission is complete [8 min read] bryl.link/7d4 #cybercartel #cybercrime

Bryley @bryleysys.bsky.social · Sep 9

"bottom line is that companies are expected to supervise their AI systems just as they would their human workforce," Atty Charles Nerko told Heller Search. From a legal standpoint, if an AI is used, its work is your organization’s work ... [7 min. read] bryl.link/4xe #AIatWork #AIlaw

Bryley @bryleysys.bsky.social · Sep 3

Traditional perimeter security misses attacks using legit software. Attackers used Chrome Remote Desktop & file-copying tools to steal data - EDR caught what firewalls couldn't ... [4 min. read] bryl.link/en8 #cybersecurity #EDR

1

Bryley @bryleysys.bsky.social · Aug 27

Happy to be recognized as the 7th largest IT Services Provider in Central Ma by Worcester Business Journal! Since 2017 Bryley's been ranked among the top in our region. We're grateful to the clients who trust us with their IT needs bryl.link/4jw #ITServices #Worcester

Bryley @bryleysys.bsky.social · Aug 22

Up Times August edition: How Bryley grew to seventh largest IT provider in Central Mass, EDR in action, lots more like “concentrated benefits and diffuse costs” as it applies to how people tend to treat security matters: bryl.link/tjm #economicsandsecurity #edr #wbj

Bryley @bryleysys.bsky.social · Aug 4

Ars Technica on the Senate-stopped 10-year moratorium to stay individual states’ regulatory power over AI ... [4 min. read] bryl.link/kzx #aimoratorium #statesright

Bryley @bryleysys.bsky.social · Jul 30

Electronic Frontier Foundation (EFF) admits to limitations in what we can do to prevent triangulating our locations if the devices are on and in our pockets, still EFF offers a guide of what we can do better our privacy [10 min. read] bryl.link/dq0 #eff #privacy

Bryley @bryleysys.bsky.social · Jul 28

Just because something worked when it was new, doesn’t mean it won’t need attention down the road. Steve Gibson points out what organizations can learn from recent telecom breaches bryl.link/6hv #patching #peaceofmindit

Bryley @bryleysys.bsky.social · Jul 25

"We very quickly learn you can only ask [a Magic 8-Ball] yes/no questions ... we're doing the same thing with chatbots. We ... tell ourselves a story as we put [input and] use that story to make sense of what comes back" Dr Emily Bender [37min] bryl.link/ek8 #magic8ball

Bryley @bryleysys.bsky.social · Jul 23

How to deal with the 'if I patch this, will it break everything?' conundrum. And the value of another perspective on an organization's security weaknesses. What we can take from Salt Typhoon's telecom breach ... [4 min read] bryl.link/jwn #lessonsfromthebreach #security

Bryley @bryleysys.bsky.social · Jul 17

Bryley's 11th MSP 501 award – an independent evaluation, so organizations can better recognize which IT companies have built something solid, worthy of their trust … [3 min. read] bryl.link/xe7 #msp501 #newenglandit #itservices

Bryley @bryleysys.bsky.social · Jul 15

July 2025 Up Times by Bryley newsletter: Bryley wins 11th MSP 501, Salt Typhoon lessons, cell phone privacy tips, Computer History Museum – linguist Emily Bender on LLMs bryl.link/lzd #MSP501

Bryley @bryleysys.bsky.social · Jul 10

2024's "Change Healthcare cyberattack was attributed to human risk. [An] employee's credentials were compromised through a phishing email [giving] attackers ... network access [to] deploy ransomware" Mimecast shares about human-error mitigation bryl.link/k46 #ransomware

Bryley @bryleysys.bsky.social · Jul 8

Google releases Gemini for under-13s. "These human-like interactions will be confusing, and potentially risky, for young children. They may believe content can be trusted ..." – Researcher Lisa Given [5 min. read] bryl.link/qka #kidsandai #cybercrime

Bryley @bryleysys.bsky.social · Jul 2

"pay a bunch of math PhDs $100/hour to come up with a lot of math problems ... [Then feed] an existing model, like GPT-4 ... using reinforcement learning techniques to tell it exactly where ... its answers [are] right or wrong ..." [12 min] bryl.link/0rl #ai

Bryley @bryleysys.bsky.social · Jul 1

"each user empties their cryptocurrency wallet into [a] tumbler. The tumbler swaps Alice's money with Bob's money ... When Alice gets her money back ... the currency she receives doesn't contain any of [her original] money" - Barracuda [10 min] bryl.link/jgg #cybercrime

Bryley @bryleysys.bsky.social · Jun 27

Matt Parker's Harvard talk covers fake security measures - like this ineffective lock. Seen security measures that are show not substance? Worth a watch. [1 hr] bryl.link/hvi #securitytheater #cybersecurity

1

Bryley @bryleysys.bsky.social · Jun 26

Bryley stopped a session-hijacking attack and @Huntress' ITDR solution was a standout help: robust identity threat detection and definite ROI ... [5 min. read] bryl.link/61t #itdr #cybersecurity

1