1. Cybercrime & Cybersecurity
1.1. Cybercrime
Payment Fraud: The latest report from the Banque de France’s Payment Security Observatory warns of a sharp increase in fraud involving tampering (+34% year-over-year), which is projected to total 516 million euros in 2025 and account for 41.6% of total payment fraud (Press release dated September 7th, 2026).
Organised crime: Europol analyses the most threatening criminal networks in the European Union and highlights their ability to capitalise swiftly on opportunities, diversify their activities and adapt to technological and economic developments. In particular, the report emphasises their increasing use of digital services, legitimate business structures and corruption to facilitate their criminal activities (Europol report of 21 August 2026)
1.2. Cybersecurity
ANSSI: The National Agency for Information System Security has announced the launch of a new initiative called REACTIV (Interministerial Response and Action to Data Breaches). It is a response to the rise in cyberattacks, particularly data breaches affecting government agencies. (Roadmap for Priority Efforts in Government Cybersecurity 2026–2027, September 7, 2026)
Cyber Resilience Act: Starting September 11, 2026, the first reporting requirements under the European Cyber Resilience Act will take effect. Manufacturers are now required to report actively exploited vulnerabilities and serious incidents to ENISA, the European Union’s central cybersecurity agency. (Regulation (EU) 2024/2847 “Cyber Resilience Act”)
AI and cybersecurity: During an internal assessment, OpenAI’s AI agents bypassed the isolation measures in their environment, exploited a zero-day vulnerability and compromised Hugging Face’s infrastructure. The incident highlights the growing capabilities of AI systems to carry out complex cyber operations (OpenAI press release, 26 August 2026).
2. Desinformation and information warfare
Disinformation: OpenAI has removed a number of ChatGPT accounts linked to a Russian disinformation campaign promoting the International Burke Institute, a fake think tank based in Israel. The accounts were using ChatGPT to generate content posted on several social media platforms to promote pro-Russian narratives (OpenAI press release, 25 August 2026).
3. Personal data and privacy
3.1. Data breaches and incidents
Electronic Invoicing: The Welyb electronic invoicing platform suffered a data breach on September 1, 2026. The data potentially stolen includes personal information as well as SIREN/SIRET numbers. This breach comes as France has just begun the transition toward widespread adoption of electronic invoicing. (Press release dated September 16, 2026)
Data Breach: Cybercriminals tricked the online banking service Revolut by impersonating a government agency, allowing them to obtain sensitive data from certain customers. The hackers are now threatening to release new files every day if the bank does not pay a ransom (Press release dated September 12, 2026)
Fingerprinting: AliExpress is reportedly using the WebAudio API to generate an inaudible sound signal and analyse the device’s response in order to create a digital fingerprint that enables it to track users without using cookies (Press release dated 24 August 2026).
3.2. Penalties and regulations
Personal data: TikTok and ByteDance have reached a $400 million settlement with the US Department of Justice to settle legal proceedings concerning the collection of data from children under the age of 13 without parental consent, in breach of COPPA. TikTok will pay $300 million immediately and a further $100 million subject to certain conditions (US Department of Justice press release of 21 August 2026).
Facebook is verifying its human users: Meta is launching Facebook Verified, a free programme enabling users aged 18 and over to obtain a badge following verification via a video selfie. This measure aims to distinguish real people from fake profiles and automated accounts, but involves the collection of additional biometric data (Press release dated 24 July 2026)
4. Digital Economy and Competition
Publication of the Code of Best Practices for the Sale of Refurbished Digital Devices on Marketplaces (2026): Signed by online marketplaces under the auspices of the Directorate General for Enterprise, the Charter outlines the legal obligations of sellers and platforms (VAT, Extended Producer Responsibility, Waste Management) and proposes voluntary commitments to strengthen transparency, compliance, and the circular economy in this sector (Press Release dated September 15, 2026).
ICANN: On August 18, 2026, Pavel Durov, CEO of Telegram, announced that he had submitted an application to ICANN to manage the .gram domain registry. ICANN is expected to make its decision in September (Press release dated August 18, 2026).
Domain names: On 12 August, ICANN closed its 2026 application round for new gTLD extensions, having received over 1,600 applications. The proposed extensions will be published following administrative review on ‘Reveal Day’ (ICANN press release of 13 August 2026)
Video games and digital sovereignty: The European Video Games Observatory proposes increasing public support for the sector, in particular by funding marketing and distribution, and regulating ‘dark patterns’ and attention-extraction models. It also calls for the development of sovereign European digital infrastructures (European Video Games Observatory report of 1 September 2026).
5. Artificial Intelligence
5.1. Copyright in the age of IA
Microsoft and AI training: Microsoft maintains that training AI models on copyright-protected books and using them does not replace the original works, thereby defending the transformative nature of such uses under ‘fair use’. This position is contested by authors, who believe that AI threatens the book market and creators’ remuneration (Press release of 4 September 2026).
Copyright: Carlsen Verlag, Marc-Uwe Kling and Astrid Henn have brought legal proceedings against OpenAI in Germany, alleging that ChatGPT has reproduced copyright-protected material from the book *Das NEINhorn* and objecting to the use of their works to train AI models (Press release from Carlsen Verlag dated 19 August 2026).
Sony Music takes legal action against Anthropic: Sony Music Publishing and Warner Chappell Music are suing Anthropic, accusing it of using thousands of copyright-protected musical works – obtained in particular through illegal downloading – to train its Claude models. The publishers are claiming up to $150,000 per infringed work and are seeking an injunction to stop these practices (Press release dated 28 August 2026)
5.2. Regulation and supervision
AI Development: A former researcher at Anthropic condemns OpenAI and Anthropic’s irresponsible race toward a self-improving superintelligence that he considers dangerous to humanity and calls for global coordination to avert a catastrophe (Press release dated September 9, 2026).
AI Development: Dario Amodei (Anthropic), Sam Altman (OpenAI), and Elon Musk (Grok) are calling for a slowdown in AI development to regain control, fearing that a swarm of misaligned agents could take control of the internet within 6 to 12 months, with a risk of more than 10% that these systems could destroy civilization (Press release dated September 12, 2026).
Hyper-manipulation and cultural sectors: The Higher Council for Literary and Artistic Property (CSPLA) has drawn up an assessment of the risks associated with AI-generated deepfakes, particularly with regard to personality rights, trust in information and copyright. The report also highlights the limitations of the current legal framework and of detection and labelling mechanisms (Progress Report of the Higher Council for Literary and Artistic Property, September 7th 2026)
AI Development: The G20 has adopted an approach that supports the development of artificial intelligence over strict regulation, thereby aligning with the United States’ policy of promoting innovation. Europe, for its part, remains cautious and emphasizes risk management, giving member states the freedom to set their own rules. (Press release from the U.S. Department of Commerce, September 3, 2026)
6. Intellectual Property and counterfeiting
6.1. Legal actions and proceedings
Anthropic Agreement with Book Rights Holders: The American AI company had reached a financial agreement with the rights holders of thousands of books following the recovery of 40,000 titles from pirate libraries such as LibGen. The problem now is identifying who actually owns the rights to each book: the author, publisher, co-author, co-publisher, etc. (Notice of Settlement Agreement, Bartz v. Anthropic, July 20, 2026)
Copyright: a proprietor of a trade mark with a reputation may prevent that mark from being used by a political party where the latter has not demonstrated that its freedom of expression takes precedence over the rights and interests of that proprietor (CJEU, September 8, 2026, Inter IKEA Systems, C-298/23)
Copyright / Social Media: The Court of Justice of the European Union (CJEU) has reiterated that a text published on a social media platform may be protected by copyright. The Court holds that a short text can constitute a “work” if it is original and reflects the author’s personal choices. The length of the text or the fact that it is published for free on the Internet therefore does not preclude its protection. (CJEU, September 3, 2026, Gândul Media Network, C-598/24)
6.2. Regulation and supervision
Fight against illegal streaming: The IFPI (International Federation of the Phonographic Industry) announces the launch of the Streaming Integrity Initiative (SII), a new industry commitment to combat streaming fraud and protect artists, songwriters, and fans from its harmful effects (Press release dated September 14, 2026)
Protection of Cultural Works: French President Emmanuel Macron and the South Korean President adopted a joint statement at the Lumière Summit on September 7 to strengthen international cooperation, protect cultural works, and combat piracy in the film and moving-image industries. (Press release dated September 7th, 2026)
7. Regulation & Justice
7.1. French Law
7.2. European Law
DSA: The European Commission has designated ChatGPT as a very large search engine (VLOSE), and Reddit and Roblox as very large online platforms (VLOP), as these services exceed the threshold of 45 million monthly users in the EU. They will be required to comply with the DSA’s enhanced obligations by January 2027, particularly regarding systemic risks and the protection of minors (European Commission press release of 31 August 2026)
CJEU: Google may be liable for its partners’ content: The CJEU has ruled that Google may lose the liability exemption applicable to hosting providers when it plays an active role in relation to the content of a creator with whom it has entered into a commercial partnership, in particular by reviewing that creator’s videos prior to the agreement. The judgment concerns the promotion of gambling on YouTube (CJEU judgment in Case C-421/24 of 16 July 2026).
