The Security Check-In · AI & the Threat Landscape: 2026 & Beyond
The Security Check-In · Ep. 02
AI & the Threat Landscape
The Security Check-In · Ep. 02 · Quarterly

AI & the Threat Landscape:
2026 & Beyond

A plain-English look at how AI reshaped cyber threats this year — and where they're headed next for your business.

Tuesday, November 10, 2026  ·  11:30 AM MST  ·  Hosted by Caldera Cybersecurity
Before we dive in

Let's make this a conversation.

No slide marathon, no acronym soup. Ask questions anytime — the best parts are the ones you bring. We'll look at the four shifts that defined 2026 — and where each one is heading:

01AI-powered scams & phishing
02Deepfakes & voice cloning
03Agentic AI — attacks that run themselves
04Fighting AI with AI
The year in review

2026 was the year AI changed the game — for both sides.

The big shift: AI stopped being just the attackers' writing assistant and became the attacker — autonomous agents that scan, exploit, and move on their own. Lures went flawless, voices got cloned, and small businesses were squarely in range.

40%
of organizations were hit by AI-enhanced attacks this year
44%
of breaches now involve ransomware — up from 32% a year ago
$3.05B
lost to business email compromise scams in a year
Close to home · New Mexico $85.6Mlost to cybercrime5,688complaints filed$2.6Mto AI-related scams
Sources: ExtraHop Global Threat Landscape Report (2026); Verizon DBIR & FBI IC3 (2025); New Mexico figures: FBI IC3 Internet Crime Report (2025).
Smarter scams
Topic 01

AI-powered scams & phishing

The typos are gone. AI writes flawless, personalized emails — perfect grammar, your logo, a vendor's real name.

The old tell — bad spelling — is dead.
The new tells are in the details: the sender, the urgency, the link. →
Let's talk
Has anyone gotten one this convincing? What tipped you off?
Look‑alike domain Manufactured urgency Link goes elsewhere
Seeing isn't believing
Topic 02

Deepfakes & voice cloning

A few seconds of audio — a voicemail, a podcast, a video — is enough to clone a voice or a face on a live call.

The setup is always the same: urgency + secrecy + a payment.
Hearing or seeing someone is no longer proof it's them.
Let's talk
How would your team verify a call from “you” asking to move money?
Incoming call
Your CEO
Mobile · now
AI‑cloned from 3 seconds of audio
✕✓
What's coming next
Topic 03

Agentic AI: attacks that run themselves

No human at the keyboard. It scans, finds a gap, exploits it, and moves to the next target — around the clock.

Finding flaws got automated — fixing them didn't. One AI flagged 23,000+ software vulnerabilities; just 75 were patched.
One attacker can now run thousands of these at once — around the clock.
Source: Anthropic vulnerability research (Project Glasswing), 2026
Let's talk
What changes for your defenses when attacks never sleep?
24/7
no human
01Scan
02Find flaw
03Exploit
04Next target
The hopeful part
Topic 04

Fighting AI with AI

Attackers now move laterally in minutes — too fast for humans alone. The answer is to put AI on defense: watching everything, all the time.

Breakout time — how fast an intruder spreads from the first machine to the rest of your network — keeps collapsing.
SIEMless, always-on AI detection triages the noise and flags the real threat in seconds — not days.
It buys your team time — it doesn't replace judgment.
Let's talk
When an intruder can move in minutes, how fast would your team notice?
Average breakout time
First foothold → moving across the network
62min
2024
48min
2025
29min
2026
Fastest observed: 27 seconds.
Source: CrowdStrike Global Threat Report (annual)
Where to start

Four moves to make now.

Turn on MFA everywhere

Still stops the vast majority of account takeovers — AI or not.

Verify money & voice requests out loud

A quick call-back defeats deepfake and invoice fraud.

Train the team on AI-era scams

The human layer is your best filter when the fakes are flawless.

Get a second set of eyes

A partner who watches this space so you don't have to.

Your turn

What's on your mind?

Questions, war stories, “is this thing actually a risk for me?” — all fair game.

Thank you.

Want to talk through where your business stands? Grab a free Caldera Coffee Break — no pressure, no pitch.

Scan to book your free Coffee Break
grow.calderacyber.com/widget/bookings/caldera-coffee-break
[email protected]  ·  (505) 975-4470
Comply · Defend · Protect — Serving New Mexico, Southern Colorado & beyond
← / → or space
1 / 10