Google to Back Unified robots.txt Standard for Blocking AI Training Data Use
Google has confirmed it will support a unified mechanism inside robots.txt that lets website owners signal their preference regarding the use of site content for training generative AI models.
The company stated in its Search Central blog that the new capability will be introduced in 2027 and is based on proposals developed by the AIPREF working group at IETF. Rather than maintaining separate directives for every AI provider, publishers will be able to set one general policy marker that participating organizations are expected to honor.
Google has indicated it will interpret an unspecified general marker in the same manner as its existing Google-Extended directive. At the same time, current per-provider settings will remain fully supported, allowing site owners to establish a broad default rule while still applying different conditions to individual companies.
For publishers, the change is expected to reduce the operational burden of managing an expanding list of AI crawlers. However, the mechanism only communicates usage preferences and does not technically block access to servers. Its effectiveness will ultimately depend on which organizations join the initiative and consistently respect the declared rules.
The proposal is still under active development and will continue to be refined through the IETF standardization process before final implementation.
Related articles
Russian CA TCI Integrates ACME Protocol to Automate TLS Certificate Issuance
The Technical Center Internet (TCI), part of RTK-COD, has deployed a production ACME server compatible with standard clients such as acme.sh, Certbot and lego. The implementation aligns the ACME state machine with the existing isolated signing infrastructure, Domain Control Validation mechanisms and mandatory Certificate Transparency logging to Yandex and VK CT logs. Certificate lifetimes are shortening industry-wide, with Web PKI limits dropping to 47 days by 2029, making manual issuance unsustainable. The service supports HTTP-01 and DNS-01 challenges, ECDSA P-256 keys and certificate revocation, while preserving the requirement for Subject/CN in CSRs and mandatory revocation reasons. A separate staging environment uses an internal CT log simulator for safe testing. Future plans include full multi-SAN support, ARI (RFC 9773) and removal of the Subject/CN restriction.
Russia Pushes Corporate Multibanking Forward with Updated Open API Standards
The Bank of Russia and AFT have advanced Open Banking standards that enable corporate clients to aggregate account data across multiple banks within a single interface. The new specifications cover consent management, security profiles based on STO BR FAPI and OpenID Connect, and account information retrieval for both individuals and legal entities. While the 2025 pilot results demonstrated practical use cases such as personal financial assistants and corporate multibanking, the updated standards effective from October 2026 still lack a unified initiation-of-transfer component. Developers must therefore combine the new account-data APIs with separate bank-specific payment initiation mechanisms. The article explores how accounting systems or independent fintech services could become the primary workspace for financial directors managing payments, payroll, and liquidity across several banks. It also examines real-world challenges including idempotency handling, status reconciliation after network failures, and the separation of application permissions from signer authority. Examples from the UK Open Banking Limited sweeping VRP statistics illustrate how pre-authorized variable recurring payments can streamline operations while highlighting the additional policy work required for corporate supplier payments in Russia.
Hacker Attacks as Force Majeure: Russian and International Courts Diverge on Liability for Cyber Incidents
In November 2026, Russian hosting provider Beget suffered a major cyber attack that disrupted websites, virtual servers, email, DNS, and control panels for thousands of clients. Beget notified customers of circumstances resembling force majeure under Article 401 of the Russian Civil Code but stopped short of claiming automatic court recognition. Russian courts have reached inconsistent rulings on whether DDoS attacks and ransomware qualify as extraordinary and unavoidable events, examining factors such as attack scale, provider preparedness, and contractual obligations. International cases from France, Norway, the United States, and the Netherlands further illustrate that outcomes depend on evidence of prior security measures, backup integrity, and specific contractual language rather than the mere occurrence of an attack. Key precedents include disputes involving Timeweb, Expeditors, and Nordlo Haugesund where liability was limited or denied despite confirmed incidents. The analysis shows that even when force majeure is rejected, clients often struggle to recover full losses due to liability caps and proof requirements. Businesses are advised to maintain independent backups, review service agreements, and document incidents thoroughly to protect against financial exposure.
US Mandates Removal of Acronis from Federal Systems by January 2027 Over Russian Ties
US federal civilian agencies have been ordered to remove all Acronis products from their information systems by January 27, 2027, with new purchases also prohibited. The directive from the Department of Homeland Security, signed on July 31 and published in the Federal Register on October 30, covers Acronis AG along with Acronis Inc. and Acronis SCS. A nearly decade-long review examined whether Russian developers still access the code and whether supply chain risks have been sufficiently reduced, though no specific evidence of backdoors was reported by sources. After 2022, Acronis suspended operations in Russia and relocated specialists, while in 2025 Swedish firm EQT acquired a controlling stake with approval from the US CFIUS committee. Despite over 75 percent of shares now held by Western investors, these measures proved insufficient for continued access to federal systems. The ban does not extend to the private US market.