It’s hard to believe 2025 is already nearing the rearview mirror. As the year winds down, SmarterMSP.com spoke with a variety of cybersecurity experts to reflect on this year’s major developments and what they anticipate as we head into 2026. Here’s what they shared.
AI’s expanding influence on cybersecurity
Dr. Anmol Agarwal, Adjunct Professor, School of Engineering and Applied Science, George Washington University
AI’s rapid evolution is reshaping cybersecurity in ways both expected and unexpected. According to Dr. Agarwal, AI agents and automation are increasingly woven into enterprise processes—and will require dedicated security strategies of their own. While security professionals have begun discussing AI-specific vulnerabilities and AI‑enabled attacks, the broader public is largely unaware.
He anticipates more sophisticated AI-driven attacks in 2026, potentially raising public awareness as incidents become harder to ignore.
Agentic AI, deepfakes, and supply chain manipulation
Bryan Sacks, Field CISO, Myriad360
Bryan Sacks highlights a sharp escalation in attack automation. He expects to see:
- Agentic AI attacks capable of executing full MITRE ATT&CK sequences and mimicking known APT behavior.
- Near-undetectable deepfakes, which will fuel executive impersonation, fraud, and misinformation—forcing enterprises to deploy stronger verification controls, such as rotating codes for high-risk actions.
- Model poisoning in open‑source AI models, a growing supply chain threat as attackers embed malicious code into widely used repositories.
Sacks also predicts that organizations will shift from denying “shadow AI” to actively enabling it securely. With non-human identities (machine accounts, service identities, autonomous agents) multiplying, companies will need better visibility and privilege governance for these emerging identities.
Identity becomes the new battleground
Chirag Shah, Global Information Security Officer, Model N
Identity is quickly becoming “the control center for everything,” says Shah—and attackers know it. As regulatory scrutiny intensifies, identity infrastructure will become a primary target.
By 2026, companies will lean more heavily into zero-trust, moving from assuming vendor safety to verifying every connection. What to expect:
- Short‑lived credentials
- Rigorous identity validation
- Real‑time system‑to‑system monitoring
Boards and regulators want proof, not promises. Vendors will be expected to demonstrate their security posture through integrity checks, software inventories, and contractual reporting obligations.
AI‑driven threats, quantum risk & rising regulation
Vincent Lomba, Chief Technical Security Officer, Alcatel‑Lucent
Lomba cautions that while AI unlocks tremendous operational benefits, it also lowers barriers for attackers. With AI tools now accessible to non‑experts, safeguards can be stripped away, enabling rapid attack development—shrinking timelines from months to days.
In 2026, he expects:
- Generative‑AI‑powered phishing, deepfakes, and autonomous intrusions to continue rising.
- AI‑driven SOCs with predictive threat modeling and automated response.
- A growing need for transparency and oversight as AI‑enhanced cyber operations mature.
Post‑quantum readiness becomes urgent
The window before quantum systems can break current encryption has shrunk by 2–3 years. Enterprises—especially in finance, government, smart cities, and critical infrastructure—will accelerate their post‑quantum cryptography (PQC) planning. Organizations will need the capability to switch between cryptographic algorithms seamlessly as this threat evolves.
A wave of new security regulations
2026 will bring substantial compliance pressure:
- DORA implementation across financial institutions.
- EU Cyber Resilience Act (CRA) deadlines, including conformity body notifications (June 2026) and mandatory vulnerability reporting (September 2026).
- Continued rollout of NIS2, with delays driven largely by cybersecurity talent shortages.
Across sectors, organizations lack the expert workforce needed to manage increasingly complex operational environments. In defense alone, an estimated 10,000–20,000 cybersecurity professionals are still needed. Universities aren’t producing enough skilled graduates—something Lomba expects will push major investment in education and workforce development in 2026.
Cybersecurity becomes a strategic imperative
Given today’s geopolitical climate, cybersecurity is no longer purely technical. Organizations must assess adversaries, identify their weakest points, and allocate resources where they matter most.
Next week, we’ll be sharing additional expert perspectives on what surprised them most about 2025—and what trends they believe will shape cybersecurity in the year ahead.
Photo: gan chaonan / Shutterstock
This post originally appeared on Smarter MSP.