AI has the potential to revolutionize the public services industry, which provides essential services such as healthcare, electricity, water, gas and telecommunications. Public service agencies face many challenges including aging infrastructure, increasing demand and environmental regulations. AI helps critical infrastructure providers like Utilities to overcome cybercrime challenges by improving efficiency and reliability and enhancing service, coverage and sustainability. Additionally, AI can help to improve cybersecurity posture by detecting and responding to cyber threats more effectively. Cybersecurity is a critical issue for infrastructure providers because they operate critical infrastructure that supports national security and public safety. A cyberattack on a utility, for example, can have widespread consequences for our daily routine.
Costs of a cybersec incident
A cybersecurity article, released with the IEEE , explains that the average cost of recovering from a cyber attack is round about $3.86 million, internationally. The article also indicates that companies need approx. 196 days to recover from a data breach. More importantly, cyberattacks on critical infrastructures are becoming more sophisticated and frequent as attackers exploit vulnerabilities in legacy systems like SCADA or use advanced techniques such as ransomware or distributed denial-of-service (DDoS) attacks.
How Blossom Sky helps to make critical infrastructures more secure
AI, powered by Databloom’s flagship product Blossom Sky, helps critical infrastructure providers by providing a federated data lakehouse, which improves their cybersecurity readiness using various techniques such as machine learning (ML), natural language processing (NLP), computer vision (CV), and reasoning.
Reasoning in AI is a logical process that involves drawing conclusions, making predictions or constructing approaches towards a particular thought using existing knowledge. It allows AI technologies to extract critical information from large data sets and use statistical inferencing in a way that approaches human cognition. These techniques enable AI systems to learn from data, understand human language, perceive visual elements, and make decisions based on logic and evidence. Federated learning can improve the training of AI models by allowing them to learn from decentralized data without the need to centralize or share that data. This can lead to better decision-making and improved reasoning in AI models. However, it’s important to note that federated learning is just one technique among many that can be used to improve AI reasoning.
AI for cybersecurity in critical infrastructures
Critical infrastructures such as power grids, water systems, healthcare infrastructures, and transportation networks are essential for the functioning of modern societies. And they are much more vulnerable to cyberattacks that can disrupt their operations and cause massive problems for law-enforcement, municipalities, cities or infrastructures, such as harbors, public transportation or airspace. To protect these infrastructures from sophisticated and evolving threats, artificial intelligence (AI) plays a key role. Not only helping to analyze large amounts of data, identify patterns and anomalies, detect malicious activities and malware, but also by providing risk assessment and decision support. AI can also enable autonomous intelligence that can act without human intervention to respond to cyberattacks in real time. By leveraging AI for cybersecurity, critical infrastructures can enhance their resilience and reliability.
We identified the most critical use-cases for AI in critical environments:
- Anomaly detection: Anomaly detection is one use case of AI for cybersecurity in critical infrastructures. AI-based systems can analyze large volumes of data from various sources to detect deviations from normal patterns of behavior. These anomalies can indicate potential cyber threats and AI systems can alert security teams when they are detected. This capability is vital for any country’s cyber strategy.
- Threat intelligence: Threat intelligence is another use case of AI for cybersecurity in critical infrastructures. AI-based systems can use natural language processing and reasoning to collect and analyze threat intelligence from various sources. This can provide valuable insights into the tactics and motivations of cyber attackers. Federated learning powered AI can use this information to identify emerging threats and recommend appropriate countermeasures.
- Incident response: Federated learning powered AI can use computer vision models and reasoning to automate some aspects of incident response. This includes triage, analysis, containment, eradication, recovery, and reporting. AI-driven systems can recognize visual elements associated with disaster incidents and use reasoning to determine the root cause and best course of action. Federated learning powered AI can also be used in disaster prevention and management.
Comply with regulatory requirements or customer expectations
There are some points to consider as a critical infrastructure operator by conducting risk assessments and ensuring that their AI systems have features that support explainability . This includes ensuring that their AI systems are transparent and can provide explanations for their decisions . Companies may also need to implement new processes and tools such as system audits and documentation to comply with more stringent AI regulations that may be on the horizon .
Federated lakehouse empowered AI has the potential to enable critical infrastructure operators, government agencies and public services like utilities to operate more efficiently, reliably, safely, and sustainably while protecting themselves from cyber threats. Federation based technology enables organizations to train AI models on decentralized data sources, like edges or remote databases, without centralizing or sharing that data. This means they can use machine learning and AI to make better decisions without sacrificing data privacy and risking violating data privacy or international data regulations. While federated learning can help with privacy concerns, is not yet ready to help with explainability challenges; but there are researches being conducted on federated learning of explainable AI models [5,6].
Data mesh and data platform abstraction are not silver bullets or one-size-fits-all solutions. They require careful planning, design, implementation, and governance. They also require a cultural shift from centralized to decentralized data ownership and collaboration. DataBloom's Virtual Data Lakehouse offers a promising vision for how organizations can harness the power of data to deliver better value for their providers, partners, and stakeholders. Be sure you undergo a brief consultation with your DataBloom AI representative to address the challenges of implementing Blossom Sky into your data strategies.
 The Impact of AI on Cybersecurity | IEEE Computer Society
 The Road to Explainable AI in GXP-Regulated Areas | Pharmaceutical Engineering (ispe.org).
 Defining Explainable AI for Requirements Analysis | SpringerLink
 AI Regulation Is Coming (hbr.org)
 EVFL: An explainable vertical federated learning for data-oriented Artificial Intelligence systems - ScienceDirect
 An Approach to Federated Learning of Explainable Fuzzy Regression Models | IEEE Conference Publication | IEEE Xplore
Blossom Sky stands for federated data lake technology, data collaboration, increased efficiency, and helping to create new insights by breaking data silos in a unified manner through a single system view. The platform is designed to adapt to a wide variety of AI algorithms and models. Blossom Sky integrates with all major data processing and streaming frameworks like Databricks, Snowflake, Cloudera, Hadoop, Teradata, Oracle, Apache Flink as well as AI systems like Tensorflow, Pandas, PyTorch.
Want to learn more? Please get in touch with us via databloom.ai/contact or write us directly: [email protected]