European Data Protection Day: three EIC-backed approaches to protecting, storing and policing data

Brussels, January 27th 2023
Summary
  • European Data Protection Day on 28 January highlights innovation in data protection and storage across EIC-backed companies.
  • Streem.ai uses real time AI to detect anomalies in large time series data sets and targets automotive, manufacturing and health sectors.
  • BioSistemika is developing DATANA, a DNA-based archival storage prototype that claims extreme density and longevity but faces cost and access challenges.
  • Illusive focuses on identity risk management to stop lateral movement in breaches and was acquired by Proofpoint in December 2022.
  • Each approach addresses different layers of data risk but also exposes practical, economic and regulatory questions for wider adoption in Europe.

Innovative approaches to handling data in Europe

European Data Protection Day is observed on 28 January and has been celebrated for over a decade. The day is an occasion to reflect on technologies and business models that shape how personal and industrial data are protected, used and preserved. Several companies that have received funding from the European Innovation Council are developing markedly different answers to data challenges. They range from real time anomaly detection to using DNA as a medium for long term archival storage and to identity focused cyber defence. Each one is promising but each also raises technical, economic or regulatory issues that will determine whether the solutions scale across European markets.

Detecting anomalies in time series: Streem.ai

Streem.ai, an EIC-backed start up, targets the problem of finding anomalies in large volumes of time series data. Anomalies in sensor streams or operational telemetry can be symptoms of defects, compromised devices or faults in clinical diagnostics. Missing those anomalies can produce downstream failures that are costly or dangerous.

How the technology works:Streem.ai deploys machine learning algorithms to identify outliers and unusual behaviour in real time across large time series datasets. The company reports that it chooses algorithms using an internal benchmarking routine that is continuously improved. The product is positioned to analyse 100 percent of available data rather than sampled subsets. Claimed benefits include up to an 85 percent reduction in analyst time and higher detection rates for previously unknown anomalies.
Intended use cases and deployment:Streem.ai is active in automotive, manufacturing, health technology and connected devices. In manufacturing the tool is used to detect inconsistencies between produced parts and to prevent faulty components progressing through production. In connected devices it can flag compromised units. In health technology it aims to surface irregular diagnostics so clinicians can offer higher quality care.
Limitations to watch for:Real time anomaly detection on high volume streams faces common challenges. Reducing false positives while catching rare, novel faults depends on model tuning and quality of training data. Adversarial noise can hide malicious activity. Integrating at edge versus cloud changes latency and privacy trade offs. Finally, vendor claims about percentage reductions in analyst time should be validated in independent, long term deployments across diverse customers.

DNA as archival medium: BioSistemika and the DATANA project

BioSistemika is a Slovenian company that provides digital solutions to laboratories and supports digital transformation. The company received EIC funding for DATANA, a patent pending project that encodes digital data in synthetic DNA. The pitch is simple. Data production is growing exponentially while traditional storage architectures and raw capacity struggle to keep pace. The company cites projections that only a fraction of produced data will be storeable within a few years unless alternatives are adopted.

Why DNA for storage?:DNA is dense and chemically stable under proper conditions. BioSistemika highlights DNA life spans measured in centuries to millennia and a theoretical information density far higher than current magnetic or solid state media. The company says DNA can hold 20 million times more data per gram than existing technologies and positions DATANA for archival roles such as data archiving, disaster recovery and custody of digital assets where long term durability matters more than immediate random access.
Technical and economic challenges:Encoding, synthesising and later reading data stored in DNA requires biochemical processes that remain expensive and relatively slow. Error rates in synthesis and sequencing must be overcome with error correcting codes and redundancy. Retrieval latency makes DNA unsuitable for frequent access workloads. Cost per byte and energy costs for synthesis remain orders of magnitude higher than tape or cold cloud storage today. Those constraints currently limit DNA storage to niche archival cases where density and longevity justify higher costs.
Company perspective and claims:BioSistemika reports DATANA is at prototype stage and that it has completed a case study with early adopters. Roswita Golčer Hratnik, CEO of BioSistemika said that DNA storage could address common pain points such as capacity and cybersecurity and listed data archiving, disaster recovery and digital asset custody as promising market areas. DATANA project manager Dr. Tomaž Karčnik emphasised that DNA storage shifts the human factor out of some protection chains and argued that DNA combines mature technological building blocks with a fundamentally different physical substrate to improve long term data safety.

Those statements underscore DNA storage prospects but also point to a need for independent validation on costs, lifecycle emissions and operational workflows. Regulators and data custodians will want clear evidence on integrity, chain of custody and the conditions required to preserve the molecule over decades.

Eliminating identity vulnerabilities: Illusive

Illusive focuses on identity risk management, a field that treats identity and privileged credentials as the primary vectors attackers exploit to move laterally inside networks. The company has developed technology designed to discover and mitigate exploitable identity vulnerabilities, stop lateral movement and provide risk aware, real time attack intelligence that accelerates triage and response.

How Illusive approaches the problem:Illusive uses deception and discovery techniques to surface credentials, misconfigurations and artifacts that an attacker could use to escalate privileges. The platform aims to be scalable, operationally simple and invisible to attackers. It provides intelligence about identity risks and attempts to automatically mitigate vulnerable configurations so that attackers cannot exploit them as entry points.
Market development and acquisition:In December 2022 Illusive signed a definitive agreement to be acquired by Proofpoint Inc. Proofpoint is a leading US cybersecurity and compliance company. Illusive founder and CEO Ofer Israeli commented that joining Proofpoint would add Illusive's approach to ITDR to a people centric security portfolio and help organisations remediate privileged identity risks that could lead to access to critical data and intellectual property.
Considerations and limits:Identity risk management is an important component of a layered cyber defence but it is not a silver bullet. Human error, shadow IT and complex enterprise workflows mean that discovery and automated mitigation can produce friction for administrators. There are also integration questions with existing endpoint, identity provider and security information platforms. The acquisition by a US vendor may offer scale and distribution but it also prompts questions about data residency, European vendor ecosystem health and strategic autonomy in cybersecurity.

Comparative summary

CompanyCore technologyEIC supportDeployment stagePrimary use casesMain risks and open questions
Streem.aiAI anomaly detection for time seriesEIC-fundedCommercial deployments in automotive, manufacturing, health and IoTFault detection, compromised device detection, irregular diagnosticsFalse positives, adversarial manipulation, integration and validation across customers
BioSistemika / DATANADNA based archival storage (prototype)EIC-fundedPrototype and early adopter case studyLong term archival, disaster recovery, digital asset custodyHigh cost, slow access, synthesis sequencing errors, energy and lifecycle impacts
IllusiveIdentity risk management and deception technologyEIC-fundedCommercial; acquired by Proofpoint in Dec 2022Stopping lateral movement, ITDR, privileged identity protectionOperational complexity, integration, vendor consolidation and data residency

Policy, funding and market implications for Europe

Funding from the European Innovation Council helps early stage companies validate and scale technologies that align with strategic EU objectives such as digital sovereignty, strong data protection and competitive industry. The three companies discussed show different pathways to impact. Detection and identity focused technologies can be adopted faster because they integrate with existing operational systems. Radical storage innovations like DNA need longer development cycles and ecosystem changes in supply chains and standards.

EU policy frameworks including the GDPR set legal guardrails for personal data processing and storage. Innovations that change where and how data are stored or analysed will have to contend with data sovereignty rules, cross border transfer restrictions and certification expectations. The fact that Illusive was acquired by a US vendor highlights a familiar pattern in cybersecurity where European scale ups are integrated into larger international incumbents. That can accelerate deployment but may also reduce the number of independent European vendors in strategic areas.

Final takeaways

EIC backed innovation is producing a range of technical responses to data protection challenges. Each solution addresses a different point in the data lifecycle and each has realistic limits. Real time anomaly detection and identity centric defences can reduce operational risk now if deployed with care. DNA storage offers a provocative long term alternative for archival needs but requires further work on cost, standards and workflows before it becomes a mainstream option. For policy makers and customers the task is to balance sensible adoption with independent validation and governance so that promising technologies translate into real improvements for data protection across Europe.