10 European startups fighting the next generation of AI cyber threats

As artificial intelligence reshapes how companies build, test and operate software, it is also transforming the cybersecurity landscape, leaving many organisations struggling to keep pace with emerging threats.

Recent research from AI Score, examining UK-based organisations across the public and private sectors, found that 31% had experienced an AI-related incident or near miss in the past two years. Among those reporting incidents, 44% said confidential or sensitive data had been entered into an AI tool.

Despite these risks, 92% of organisations are already using AI beyond the pilot stage, while significant gaps remain in their ability to identify, monitor and control its use.

AI-powered phishing, automated attacks and increasingly sophisticated impersonation techniques are creating new cybersecurity challenges. At the same time, the rapid adoption of AI agents, autonomous systems and AI-assisted software development is introducing additional vulnerabilities that traditional security tools may struggle to address.

Across Europe, a growing number of startups are developing solutions to tackle these emerging threats. Some are building tools to protect AI agents and connected systems from exploitation, while others are leveraging AI to automate penetration testing, improve threat detection and accelerate security investigations.

From securing AI-powered systems to using the technology to strengthen cyber defences, these 10 European startups are tackling the generation of cybersecurity challenges as both attackers and defenders increasingly turn to artificial intelligence.

Founded in 2020, UK cybersecurity startup Cytix is tackling the risks created by AI-driven software change. In August 2026, the company raised €6 million in a Series A to tackle the problems created by rapid software change.

The company argues that AI-assisted development, agentic workflows and continuous delivery are outpacing traditional security models. Its platform is designed to help organisations manage security risk as software changes more quickly and more frequently.

Founded in 2020, French startup Escape is developing an AI-powered offensive security engineering platform.

In March 2026, the company raised €15.4 million in a Series A round led by Balderton. Its work is part of a broader shift towards continuous, automated offensive security, where organisations use AI-assisted testing to identify vulnerabilities before attackers can exploit them.

Founded in 2018, Italian embedded cybersecurity company Exein is developing technology to protect connected devices, autonomous systems and physical AI infrastructure against emerging cyber threats.

The company raised €234 million , reaching unicorn status as it expands its approach to “ Physical AI cybersecurity “.

Unlike traditional security solutions that protect networks from the outside, Exein embeds protection directly into devices and operating environments. Its runtime security technology continuously monitors system behaviour, detects suspicious activity and can block malicious actions before they execute.

Founded in 2026, Fleuret AI is automating penetration testing with AI agents.

The Paris-based cybersecurity startup closed a €4 million pre-Seed round in October 2026 to hire AI, software engineering and offensive security talent, as well as accelerate platform development. By combining offensive security expertise with agentic AI, Fleuret is targeting more continuous and accessible penetration testing.

Hadrian (UK / Netherlands)

Founded in Amsterdam in 2021, Hadrian is building an agentic AI offensive security platform designed to continuously discover, validate and prioritise exposures across organisations’ digital attack surfaces.

In October 2026, the London-Amsterdam startup raised €35.68 million to expand across EMEA and the US, while deepening investment in engineering and research. Its platform sits at the intersection of offensive security, automation and AI-driven attack-surface management.

Founded in 2022, Barcelona-based NeuralTrust focuses on protecting enterprise AI applications and autonomous agents against emerging security threats.

NeuralTrust’s platform monitors interactions between AI agents, users and connected tools, applying security controls in real time to detect and prevent potentially harmful behaviour.

In June 2026, the startup announced a €17.2 million Seed round , which is described as the largest cybersecurity Seed financing raised by an EU company to date.

As businesses increasingly rely on AI agents to access information and carry out tasks autonomously, NeuralTrust aims to help organisations prevent attackers from manipulating these systems or exploiting their access to sensitive enterprise environments.

Founded in 2023, Paris-based Qevlar AI is using agentic AI to automate Security Operations Centre investigations.

The startup raised €25.8 million in March 2026 for its platform, which aims to support security teams by automating and deepening cyber investigations. As SOC (Security Operations Centre) teams face rising alert volumes and increasingly automated threats, Qevlar AI is positioning itself around faster, AI-assisted incident analysis.

QuoIntelligence (Germany)

Founded in 2020, Frankfurt-based QuoIntelligence provides EU-compliant threat intelligence, combining an AI-powered platform with human analysts.

The company closed a €7.3 million Series A in April 2026 to scale its work with organisations facing cyber risk, regulatory pressure and supply-chain oversight.

With European companies adapting to frameworks such as NIS2, which strengthens cybersecurity requirements across critical sectors, and DORA, which focuses on digital operational resilience in the financial sector, QuoIntelligence is focused on contextualised, sector-specific intelligence.

Founded in 2025, London-based Trent AI emerged from stealth in April 2026 with an €11 million Seed round to develop a layered cybersecurity platform built for the agentic era.

The startup is addressing the security challenges created by autonomous AI agents, which can interact with enterprise systems, access sensitive information and execute tasks with limited human supervision. These capabilities introduce new attack surfaces, including the potential misuse of permissions, compromised workflows and malicious manipulation of agent behaviour.

Zepo Intelligence (Spain)

Founded in 2021, Spain-based Zepo Intelligence is taking on AI-driven human-targeted cyber threats by enabling organisations to simulate real-world, AI-driven social engineering attacks such deepfakes, voice calls, cross-platform messages, or personalised phishing, and identify vulnerabilities

The startup raised €12.8 million in January of this year to protect workplaces from attacks such as impersonation, social engineering and other forms of human-layer compromise. Zepo Intelligence describes itself as an agentic social intelligence platform for workspace security.

Together, these companies show how Europe’s cybersecurity sector is evolving beyond traditional perimeter defence. The focus is not only on countering AI-powered cyberattacks, but also on securing autonomous AI agents, identifying vulnerabilities before they can be exploited, automating security operations and protecting the connected systems that increasingly underpin the digital economy.

BTW : Are you a corporate or investor looking to invest or acquire exciting startups in specific markets? Check out our Startups Sourcing Service !