Artificial intelligence (AI) is undoubtedly a game-changing milestone for the technology industry. While the full potential is only gradually becoming apparent, it is already becoming apparent that AI applications require intensive use of data. The result: effective and responsible data management becomes a decisive factor for a large number of use cases.
Companies are already facing the challenge of coping with the enormous flood of data in the multi-cloud world. Most conventional data management concepts prove to be insufficient due to a lack of scalability, speed and transparency. To increase efficiency, companies need to rethink their business processes and analyze how to integrate AI into their data management strategy. Implemented correctly, the AI strategy can become a regular part of a company’s overall data management.
“As in numerous other fields, the use of AI enables increasingly sophisticated automated processes in data management. With the help of AI, decisions can be made, warnings of potential situations can be issued and solutions can be proposed based on human behavior and human decisions,” explains Ralf Baumann, Country Manager Germany at Veritas Technologies. “These operations are performed with an efficiency that surpasses conventional technology and human thinking. In parallel to the application of AI, machine learning technologies are used to automatically enrich the existing information stocks. This enriched data forms the basis on which the AI makes its decisions.”
In terms of data management, AI is proving to be particularly relevant to three main activities:
1. Detect anomalies
In complex, hybrid, multi-cloud environments, the use of autonomous, AI-supported data management (ADM) has emerged as a proven best practice. With AI-based malware scanning and anomaly detection, companies can, for example, optimize the management of their data and automate protection against cyber threats such as ransomware. The integration of AI also enables the automation of data management processes, reducing manual intervention and thus increasing operational efficiencies, service levels and data archiving – with the result of accelerated decision making.
2. Predictive Maintenance
Based on knowledge about hardware and software systems as well as ongoing system monitoring, AI technologies are able to predict potential disruptions or failures at an early stage. You can take corrective or repair measures yourself, or alternatively submit these options for action to the IT teams.
3. Governance and Compliance
AI has the ability to automatically classify data based on its content. It can use filters or sophisticated automations to make decisions about the classification, storage and archiving of data – including geolocation, access rights, security levels and the like. Furthermore, AI is able to sort real-time data and user information based on recognized patterns. The combination of all three activities marks a crucial step towards fully automated data management. However, despite significant advances, AI technologies are not yet mature in all aspects. False positives continue to pose a challenge when integrating AI into data management solutions. Deployed systems must find a balance between automation and human interaction. Human intervention is necessary to at least verify the AI’s selections and decisions, even if this means that its role is limited to that of a warning. In short: Complete automation of decisions should be avoided in order to prevent potentially costly wrong decisions. As such, it is critical that IT teams use AI in a deliberate and controlled manner to ensure it is aligned with business goals.
Autonomous solutions through artificial intelligence
AI-supported, autonomous data management is a goal worth striving for. The integration of AI-supported anomaly detection, for example, has proven itself in more and more companies. Last but not least, AI is an integral part of various data governance and compliance products such as Alta Archive, Alta Discovery, Alta Information Classifier and especially Alta Data Insight. In this context, it is of great importance to further develop the range of cloud-based solutions in the direction of AI integration.
Training for products that contain AI elements also enables partners and customers to make optimal use of the range of functions, both technically and economically. For example, they can be managed by a training department that offers face-to-face training, video training or online training depending on the needs, availability and preference of customers or partners. The aim is always to offer relevant measures that also correspond to the strategic goals of the respective interested party.
How can companies protect themselves against cyber threats?
The main threat posed by cyber threats to AI technologies is that the source code or data used to train the AI is compromised. In this case, the decisions made by the AI or suggestions to the IT team may be changed and the AI’s ability to complete its original task may be affected. Manipulations like this open up opportunities for various types of cyberattacks that can affect all systems.
“To ward off such attacks, it is essential to effectively secure the AI’s training and reference data, the underlying code and the supply chain – a challenge, especially when both are based on open source,” said Baumann. “Another security mechanism is to limit the AI’s autonomy in its decisions and work closely with IT and business teams to determine which decisions and automations require human review.”
Source: Veritas Technologies LLC, Sep 5, 2023
https://www.veritas.com/