Sources on the IDScan.net breach, for anyone who wants to go deeper:
pymnts.com/cybersecurity/ide…
therecord.media/idscan-data-…
38
Your AI Agent's Login Is a Photocopied ID.
Every "AI agent" your team has deployed this year is probably logging in exactly the way a script did in 2015: one shared API key, pasted into a config file, that never expires and proves nothing about which specific agent is acting, on whose behalf, or for what task.
NIST just called this out directly.
In February, their new AI Agent Standards Initiative warned that agentic AI rollouts are prioritizing speed over security, reverting to the shared-credential pattern that causes most real-world breaches. Now it's just an autonomous, faster-acting actor holding the keys.
Here's the comparison that made it click for me:
Today's credential: A photocopied ID card. Whoever holds the copy gets the same access, forever, and there's no way to tell instances apart.
Tomorrow's credential: A limited power of attorney. "You may book this one flight, under $500, expiring in one hour, on my authority alone." Signed. Scoped. Revocable.
Deutsche Telekom and Vodafone just showed what that looks like in production. Their TM Forum Catalyst had a compliance AI agent read a new regulation, draft a network-access policy, route it to a human for approval inside a proper identity layer, and generate a tamper-evident audit trail automatically. What used to take engineers weeks now takes a day.
FIDO is building this into its roadmap, too. Authenticate 2026 in October now has a dedicated track for non-human authentication, built specifically around agentic AI and MCP.
A standards agency, a group of telecom operators, and the authentication industry itself are all arriving at the same conclusion: Identity is becoming the control layer for whether an autonomous agent can be trusted at all.
If you're running agent-based workflows right now, can you actually answer "which agent instance did this" when something goes wrong? Curious how other teams are solving it.
#IdentitySecurity #AgenticAI #NonHumanIdentity #DigitalTrust #CyberSecurity
🤖 Made with AI
1
1
66
Sources for those who want to dig into the NIST paper, the telco Catalyst, and FIDO's Authenticate 2026 track:
biometricupdate.com/202608/n…
biometricupdate.com/202608/l…
biometricupdate.com/202608/a…
1
21
The threat landscape has officially crossed a threshold: we are no longer just defending against human operators. We are defending against autonomous agents.Three separate reports published this month paint a clear picture of where cybercrime and financial fraud are heading:
THE SHIFT TO AGENTIC AI: Research published this week on the Microsoft Cloud Blog highlights that threat actors are moving past isolated deepfakes and basic phishing. They are increasingly deploying "Agentic AI"—autonomous systems capable of executing multi-step fraud campaigns at scale. In fact, 75% of anti-financial-crime professionals now flag malicious generative AI as a critical external risk.
"PROOF OF CONCEPT" IN THE WILD: This is not a theoretical threat. NCC Group’s newly released July Threat Intelligence Report confirms the emergence of "JADEPUFFER," which is assessed to be the first known fully autonomous, end-to-end AI-driven attack agent. It has demonstrated the ability to execute an attack from initial compromise through to extortion without any human instruction.
THE COLLAPSE OF TIME: Because of this automation, the speed of compromise is accelerating exponentially. Palo Alto Networks' Unit 42 recently noted that AI-enabled threat actors have fundamentally collapsed the window between breach and exfiltration, with attacks moving up to four times faster over the past year and the CVE-to-exploit window dropping to roughly 10 hours.
At the Mastercard Global Intelligence and Cyber Centre of Excellence here in Vancouver, our focus is squarely on anticipating this paradigm shift. You cannot fight automated, machine-speed threats with legacy, human-speed defenses. Securing the global digital economy requires integrating advanced threat intelligence that can proactively anticipate and neutralize these AI-driven campaigns before they execute.
For those tracking this evolution, the full reports from Microsoft Research, NCC Group, and Unit 42 are essential reading—feel free to drop a comment or DM me and I will happily send you the direct links.
#ArtificialIntelligence #Cybersecurity #ThreatIntelligence #FraudPrevention #Fintech #Mastercard #EnterpriseSecurity
1
149
Vancouver has officially been ranked Canada's #2 city for tech talent and #9 across all of North America. It’s amazing to see our local ecosystem recognized with such a strong ranking. Proud to see Vancouver continuing to hold its own among the top tech markets in the world. – at Vancouver, British Columbia
3
33
Great week at the Mastercard IoT Lab at @cic_unb. For 3+ years, this partnership has allowed us to tackle real challenges in cybersecurity, trusted AI, and digital identity. Proud to see Canada’s next-gen talent building innovations that will make a global impact. 🇨🇦🚀
2
43
Just built a private "SecondBrain" over my own documents - tax notices, contracts, bills, scans - on a local LLM that runs 100% on my Mac Mini. No cloud, monthly bills, no data leaving my house. DeepSeek-powered RAG, OCR for scans, voice questions. A librarian who never sleeps 🧠
🤖 Made with AI
2
64
Finding and attracting web dev clients looks entirely different today. We're using autonomous agents to scout brick-and-mortars, rigorously qualify leads, and deliver live proof-of-concepts instantly. The hustle has changed.
this is f**king dangerous
A guy built a system of 7 Claude agents on his MacBook.
No assistant. No sales team. No office.
Every day it scans Google Maps across 3 cities, finds small businesses with no website or one from 2014, builds a landing page mockup, renders a 10-second video of it, and sends a personalized cold message — before he wakes up.
47 clients a month. $400 each.
$18,800/month. $480 in API costs.
Traditional web agencies run 8-person teams for the same order flow.
He runs it alone from a MacBook and an iPhone.
When a positive reply comes in while he's in a taxi, his Mobile agent books the Zoom call. He taps "approve" and joins 10 minutes later.
The only time the system wakes him is when a deal breaks $3,000 or the reply rate drops below 12%.
Everything else runs without him.
Here's the complete playbook for building such $10K/month passive income machine with AI ↓
1
67
I'm deeply thankful to the University of New Brunswick. It was an honor to discuss the evolving landscape of GenAI in cybersecurity on such an insightful panel. Our conversations delved into GenAI, emerging threats, role of partnerships, & the importance of diversity in the field
4
89
Mansur Mirani retweeted
Innovation knows no bounds in 🇨🇦. Have a look at @MastercardCA's Tech Hub in Vancouver! We remain committed to supporting companies who choose to do business in Canada and bring some of the best minds, fostering collaboration & propelling groundbreaking advancements here at home! – at Vancouver, British Columbia
Mansur Mirani retweeted
I thought squid game was a Netflix documentary about sea creatures.
This dunya will always deceive you.
Mansur Mirani retweeted
We ask you to join us in sending this letter to Board Chair Cordova, and CC’d Minister Leece and show your support to renaming the school after Hodan Nalayeh (@HodanTV)
nccm.ca/hodan/


























