Publication date: October 23, 2025
Any activity that involves the processing of personally identifiable information is considered personal data processing. However, it’s important to remember that GDPR regulations do not apply to activities undertaken solely for personal or household purposes.
Publication date: October 23, 2025
Recently, administrative courts have been considering the possibility of imposing VAT obligations on individuals and companies providing services via the SaaS model. This article provides an analysis of the nature of the SaaS model, its advantages and disadvantages, the legal obligations associated with it for both users and service providers and the latest administrative court case law related to this model.
Publication date: October 23, 2025
KIEŁTYKA GŁADKOWSKI KG LEGAL participated in the conference “Longevity Biotech: Investing in a healthier future”. As part of its life science specialisation, our biotech and pharma lawyers offer comprehensive legal services to its clients in many areas of law, including extensive legal services to companies operating in the longevity biotech sector and biotech companies engaged in life extension and human health. Within this specialism our law firm offers assistance in respect of trademark protection and patent issues, regulatory law, certification, contractual, and litigation and arbitration issues. At KG Legal, we know that a key role in the biotechnology sector is innovation and responsiveness to changing market needs. Therefore, we offer our clients a proactive approach, flexibility and efficiency in action. All these qualities of our firm make us recognized as one of the leading law firms in Poland and the world, especially in the field of cross-border specialization and knowledge.
Publication date: October 23, 2025
DATASETS AND SYNTHETIC DATA
Creating and developing an AI model requires an unimaginable amount of data. Once input, the model analyzes the information, performs calculations, and draws conclusions based on this data that informs its future operations. Comparing this process to that of humans, one could say that AI “learns” in this way. AI systems are trained on numerous examples and draw model patterns from them, allowing them to predict correct solutions. This process is called “AI training”. Data is currently so expensive and difficult to access that it is estimated that it may be in short supply by 2032. The answer to these problems is synthetic data. This data is generated by the AI itself, which uses parameters from real-world data and randomly generates subsequent scenarios. These scenarios are designed to faithfully reproduce the properties, complexities, and relationships observed in the original data from which they were generated. There are certain risks involved. First of all, it is about creating synthetic data based on erroneous assumptions from real data – with the continuous introduction of new real data, previous errors can be corrected and artificial intelligence will be able to “unlearn” them, whereas if the first synthetic data generated is “contaminated”, each subsequent one will also contain erroneous information. The undoubted advantage of synthetic data is that it can be used to generate subsequent scenarios that may not actually occur at all or very rarely. This is used in industries such as automotive (for simulating traffic scenarios), finance (detecting fraud), and healthcare (detecting and treating rare conditions).
Publication date: October 23, 2025
The Polish Ministry of Digital Affairs is establishing a new sectoral cybersecurity incident response team – CSIRT Cyfra. The formal establishment of CSIRT Cyfra is planned for April 2026, with full operational readiness for June 2026. The unit will be responsible for protecting digital infrastructure, and its tasks will include monitoring threats, rapidly responding to attacks, and providing technical support to institutions using digital services.
