Generative AI and cybersecurity: What Sophos experts expect in 2026

AI has dominated cybersecurity headlines for years, but as we enter 2026, the conversation is shifting from hype to hard realities. Across incident response, threat intelligence, and security operations, Sophos experts see clearer signals of where AI is truly making an impact. For IT teams already stretched thin, this isn’t theoretical — it’s reshaping daily decisions.

Defenders are having to deal with the speed and unpredictability of AI adoption inside their organizations. Sophos survey data shows IT leaders are increasingly worried about unmanaged use, data exposure, and how AI-enabled tools can amplify small mistakes. The question is no longer whether to use AI, but how to apply it responsibly to reduce noise and operational strain.

Industry signals reinforce this shift. The Gartner® Hype Cycle™  for Security Operations, in our assessment,  shows AI-powered capabilities — such as AI SOC agents, predictive modeling, and cybersecurity assistants — moving out of pure experimentation and toward more practical, augmentation-focused use cases. The emphasis is less on “autonomous AI” replacing humans, and more on AI accelerating triage, improving prioritization, and helping overstretched teams act faster and with greater confidence.

Looking ahead, Sophos experts expect AI’s real impact to be measured by outcomes, not novelty — a theme reflected in their predictions for the year ahead.

DPRK IT workers extend the use of AI for fraudulent employment 

“North Korean IT workers could use Agentic AI to enhance the survivability of their fake personas, improve the responsiveness to remote requests, and conduct remote taskings more effectively.” - Rafe Piling, Director of Threat Intelligence, Sophos X-Ops Counter Threat Unit

AI will supercharge threat actor scale and sophistication 

 “In 2026, attackers will continue to use AI as a force multiplier. AI will make it easier to weaponize known vulnerabilities, orchestrate attack campaigns, lowering the barrier for basic hacking and enabling broad, rapid exploitation across the internet. 

Payloads will be customized faster than ever, and social engineering will become increasingly tailored, including phishing that reflects open-source knowledge of individual targets. Deepfake audio and video will make BEC campaigns more convincing and far more credible, making them easier to succumb to.

AI will shift the balance of power by helping even low-skill threat actors operate with speed and precision once reserved for more experienced threat actors.” - John Peterson, Chief Development Officer 

The real attack surface: Your AI application 

“We’re likely to see major breaches from prompt injection attacks within the next year. For years, security teams have worked to shrink their internet footprint, knowing that anything exposed increases risk. 

Firewalls, VPNs, and ZTNA all aimed to reduce that exposure. Now, almost overnight, we’ve created a new attack surface: rapidly deployed AI applications. Many are internet-facing, often unauthenticated, and connected to data many businesses would consider sensitive or confidential. Even more concerning, these applications are being granted the ability to take actions on behalf of the organization. 

The speed of AI adoption is driving huge efficiency gains, but unless organizations slow down and assess these risks, they’ll reopen exposures we’ve spent decades trying to close.” - Tom Gorup, Vice President of SOC Operations

The next insider is your AI

“Organizations are racing to deploy LLMs and agents, even the ones they approve internally. By feeding these tools massive volumes of corporate data, they’re creating a new class of insider threat. When that data leaks, who’s responsible? Is the AI an ‘employee,’ and who carries the liability when it goes rogue, is compromised, or misconfigured?” - Chris O’Brien, Vice President of Security Operations

Customers will turn to channel partners to resolve the AI vs. Security priority gap

“In 2026, customers will wrestle with the tension between AI investment vs. core security investment vs. broader IT investments. They’ll be asking: ‘How much time, money and resources should I spend on AI?’ ‘Where do AI and security overlap?’ ‘What are the initial use cases that I trust AI to do?’ 

While SMBs and enterprises will answer those questions very differently, both will rely on vendors and partners as their trusted advisor. With more tools flooding the market and budgets under pressure, partners who help customers make sense of the noise and rationalize their stack will prevail.

The channel’s role is no longer about sourcing tools; it’s about guiding customers through complex decisions around risk, budget, security and AI strategy trade-offs across the entire customer lifecycle.” - Chris Bell, Senior Vice President of Global Channel, Alliances and Corporate Development

Trusting MDR when the analysts are AI 

“In 2026, the Managed Detection and Response (MDR) market will reach an inflection point. The line between a managed detection service and an AI-driven tool will blur until customers can no longer tell which they’re buying. Vendors will market software as full MDR offerings, leaning on AI to compensate for limited human depth.

That shift will create real challenges: customers won’t know where human judgment ends, what’s automated, or who is actually monitoring their environment around the clock. Trust will become harder to calibrate when the ‘team’ behind your security is mostly code. As this ambiguity grows, buyers will struggle to evaluate competence, accountability, and reliability, forcing the industry to confront whether MDR is still a service, a tool, or something in between.” - Gorup

Gartner, Hype Cycle for Security Operations, By Jonathan Nunez, Darren Livingstone, 23 June 2025.

Gartner and Hype Cycle are trademarks of Gartner, Inc. and/or its affiliates.

Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.