The Competition Commission of India (CCI) regulator dismissed a complaint filed against the company Eternal on Thursday. The commission ruled that there was no initial basis for an accusation of abuse of dominant position regarding platform fees and pricing on the Zomato food delivery application.
The case arose after an individual consumer approached the regulator. This person claimed that a dish purchased through Zomato cost 198 rupees, whereas the price when buying directly from the same restaurant was only 105 rupees. The complainant pointed out that the price difference was due to inflated menu prices, delivery charges, platform commissions, and taxes.
Furthermore, the complaint stated that restaurants were forced to increase menu prices because of the commissions charged by Zomato. The applicant alleged that this violated the provisions of Sections 3 and 4 of the Competition Act, 2002. These sections pertain to anti-competitive agreements and abuse of dominant position, respectively.
Dismissing these allegations, the commission noted that online food delivery includes additional services such as delivery and platform services, making it inappropriate to compare prices with direct purchases at restaurants. CCI stated in its order that platforms charge consumers for delivery and platform services, while restaurants may pass on commissions through menu prices.
The regulator also stated that consumers who opt for online food delivery consciously pay extra charges for the convenience of this service, and that the business models of restaurants and online delivery platforms differ. Regarding the allegation of 'drip pricing,' CCI emphasized that additional charges, including platform commissions, delivery fees, and taxes, relate to provided supplementary services, and the consumer retains the option to accept or reject the order before the final checkout stage. The antitrust authority stated: 'Thus, drip pricing does not create any competition issues.'
Consequently, the regulator concluded that there was no prima facie case of violation of the provisions of Sections 3 or 4 of the Act against the opposing party (Eternal, formerly Zomato Ltd).
Google has introduced the 'video selfie for sign-in' feature—an account recovery tool that allows users to verify their identity via a short video selfie if they lose access to their password, phone, or other login methods.
This functionality is part of Google's latest efforts to implement biometric authentication alongside the use of passkeys and multi-factor authentication (MFA). Although the company claims this feature will improve convenience and recovery security, its launch comes amid a rapid rise in deepfake fraud generated by artificial intelligence (AI).
Google positions the 'video selfie for sign-in' as an additional recovery method, not a complete replacement for passwords or passkeys. The process begins with the user recording a short reference video selfie performing specific movements, such as turning their head, so that Google's systems can capture the face from different angles. Upon subsequent loss of account access, the user can record a new video, which is then compared to the original for identity verification.
According to the company, the verification process includes several security measures, such as AI-based facial recognition, liveness detection, motion control, and detection of suspicious login attempts. These measures are intended to reduce the risk of impersonation using photographs, replayed videos, or AI-generated deepfakes.
Google states that the reference video is encrypted during storage, is collected only with user consent, and can be deleted at any time through account settings. By default, the video is used exclusively for account recovery unless the user explicitly agrees to additional use cases. The company emphasizes that the 'video selfie for sign-in' complements existing protection measures, such as passkeys, recovery contact details, and two-step verification.
Nevertheless, cybersecurity specialists urge caution, pointing to the potential misuse of facial data. Purshottam Bhatia, Head of Consumer Business in South Asia at Kaspersky, noted that simple fraudulent attempts are unlikely to succeed against sophisticated deepfake detection systems, but attacks targeting high-value accounts remain a problem.
He added that fraudulent websites often request access to the device camera under the guise of account verification. The obtained facial data can subsequently be used for unauthorized account access or sold on the dark web. Experts agree that no single authentication method is sufficient protection against modern threats; instead, systems increasingly rely on multiple independent verification methods.
Passkeys, based on FIDO Alliance standards, are considered among the most reliable authentication methods because they eliminate traditional passwords and use cryptographic keys securely stored on trusted devices. Two-Factor Authentication (2FA) adds another layer of protection by requiring users to confirm their identity via a separate device or authenticator app.
The 'video selfie for sign-in' serves a different purpose: it primarily solves one of the biggest user problems—regaining access after losing passwords or trusted devices—rather than authenticating every login. For users who frequently change or lose phones, this feature can reduce the risk of permanent account lockout while providing higher protection than traditional recovery questions.
Experts warn that one of the biggest risks associated with facial biometric authentication is its irreversibility. Amit Jaju from the consulting firm Ankura stated: 'A password can be reset after compromise; a face cannot be changed. If a face template, video selfie, or related identification data is stolen, it can create a permanent risk of fraud and privacy for the individual.'
At the core of Google's new system is AI-based liveness detection. Unlike standard facial recognition, which simply compares two images, liveness detection attempts to establish whether a person is present in front of the camera at that moment. The head movements specified by Google are designed to prevent attackers from simply presenting a photograph or replaying a recorded video.
The company claims to combine facial matching with multiple security signals and standard account protection systems to detect suspicious recovery attempts and impersonation. However, the effectiveness of these systems will depend on how quickly their development keeps pace with generative AI. Modern deepfake tools are already capable of creating convincing facial movements, realistic eye blinking, and synchronized speech, which complicates the detection of impersonation attempts.
Google's announcement coincides with a sharp increase in AI-driven fraud. According to a 2026 report by pi-labs cited by Business Standard, the volume of deepfake content has increased by 900 percent in recent years. The report noted that over 90 percent of explicit deepfakes target women, and 65 percent of Indian organizations reported deepfake-based attacks in 2026.
Furthermore, over 5,000 face-swapping applications and more than 1,000 voice cloning tools have become publicly available, significantly lowering the barrier for cybercriminals. The rise in reports of cybercrimes involving women, from approximately 50,000 cases in 2024 to nearly 80,000 by 2026, reflects the growing abuse of AI for identity manipulation and digital fraud.
Beyond security concerns, there is also the issue of privacy. Unlike passwords, biometric identifiers cannot simply be reset after a leak. If facial data is compromised, users cannot replace their face in the same way they can change a password. Google attempts to address these concerns by stating that video selfies are recorded only with user consent, are encrypted during storage, are used only for account recovery unless the user agrees otherwise, and can be deleted at any time.
According to Bhatia, every collected video selfie represents an ongoing commitment, not a check that concludes after verification. Kaspersky's 2025 phishing study also showed that attackers are moving beyond passwords and increasingly targeting biometric data, as well as electronic and handwritten signatures. He emphasized: 'None of this data can be reset like a password. Companies that collect it are effectively taking guardianship over something the user cannot get back if it leaks, so responsibility does not end when verification is complete. It continues as long as the data is stored.'
Google is not the first technology company to use biometrics for authentication. Apple uses Face ID, which performs facial recognition primarily on the device, utilizing the Secure Enclave for local processing and protection of biometric information, rather than storing the face image in the cloud. Microsoft supports biometric authentication through Windows Hello, which similarly performs verification on compatible devices without centralized storage of user face templates.
Financial institutions are also increasingly adopting selfie verification for remote registration and KYC (Know Your Customer) processes. Many banks combine facial recognition with document verification and liveness detection to reduce fraud when opening accounts or conducting other high-risk transactions. Google's implementation differs in that it functions primarily as a recovery method for one of the world's largest digital identity platforms, as a Google account grants access to Gmail, Photos, Drive, payments, and numerous third-party services.
Google's video selfie feature reflects a broader trend in the technology industry—the shift towards biometric authentication to strengthen account security and simplify recovery. For users, this can reduce the risk of permanent account lockout and add a protective layer on top of passwords and emails. However, according to cybersecurity experts, no biometric system is absolutely secure, as deepfakes and AI-driven attacks are constantly evolving. The effectiveness of selfie verification will depend on how quickly AI-based liveness detection can keep up with increasingly complex threats.
Biometric data also creates unique privacy and security risks because, unlike passwords, they cannot be changed after being breached. Experts insist that strong encryption, limited data retention, user consent, and transparency will be critical to gaining trust in such systems. Thus, the 'video selfie for sign-in' feature should be viewed as an additional layer of security, not a standalone solution, and its long-term success will depend on the company's ability to continuously enhance defense against evolving AI-driven fraud.
Previously, Chinese memory chip manufacturers were seen only as secondary players compared to global industry leaders. This situation has undergone a significant change. As reported by Reuters, companies CXMT and YMTC have begun to exert influence over pricing, secure high-value contracts, and gain a foothold in one of the most competitive segments of the semiconductor industry.
This advancement strengthens the technological rivalry between China and the United States while intensifying pressure on established market competitors.
For many years, memory chip production involved operating with low profit margins. Both ChangXin Memory Technologies (CXMT) and Yangtze Memory Technologies Corp (YMTC) had to rely on public funds and accumulate losses while attempting to compete with foreign producers.
The scenario was altered by the exponential growth of data centers focused on artificial intelligence. The demand for memory used in smartphones, computers, and servers skyrocketed, elevating the status of these components to strategically important products.
In this context, CXMT has demonstrated its new relevance in the market. Reuters reported that the company rejected requests from Huawei aimed at price reductions and prevented engineers linked to the supplier SiCarrier from returning to the factory's R&D department after a disagreement.
Currently, in addition to being able to choose their clients, Chinese manufacturers are able to charge prices higher than those practiced by companies like Samsung and SK Hynix on certain occasions. Recent milestones include:
Financial indicators highlight the magnitude of this transformation. CXMT debuted on the Shanghai Stock Exchange after an IPO that raised US$ 8.6 billion (about R$ 43.7 billion). In the first quarter, the company reported revenue of US$ 7.5 billion (approximately R$ 38.1 billion), representing a 719% year-over-year increase.
YMTC follows a similar trajectory. Sources cited by Reuters indicate that executives advocate for a valuation of 1 trillion yuan, which corresponds to US$ 148 billion (approximately R$ 751.8 billion). Both companies receive support from the state fund Big Fund and are considered vital to China's goals of achieving technological autonomy.
However, not all aspects have changed. CXMT and YMTC still depend on lithography machines supplied by ASML and remain without access to the most modern equipment from the Dutch manufacturer, which has been blocked for China since 2019.
Ray Wang, an analyst at SemiAnalysis focusing on memory and artificial intelligence supply chains, summarizes the main obstacle: 'If more restrictions are imposed on lithography equipment, this will be the biggest challenge for Chinese memory manufacturers.'
Despite these restrictions, optimism persists. In 2024, Chen Nanxiang, president of YMTC, stated that 'although the industry has not yet achieved explosive growth, that day will come in three to five years.' The succession of contracts, investments, and increased production suggests that this expectation is beginning to materialize.
The new production and office building is located in Dubrovčani, a village northwest of Zagreb, which is characterized by a rural and natural landscape. The site where the object is being built is in an area that is rapidly turning into an industrial complex and small business zone, and was originally used as a landfill for construction waste.
After remediation work on the landfill, the site was restored, leading to the formation of a new topography and a new natural appearance. Following the natural slope of the terrain, three hills were formed on the site, which defined the central space and pedestrian routes. Technical, production, and public areas are located within these hills and oriented towards the pedestrian path connecting them.
The horizontal disc serves as an anchor in this unstable context. A simple ground floor with a square plan, organized around a central circular atrium, sits atop the three green hills. The building is simultaneously neutral and specific, drawing the surrounding natural landscape inside and creating unexpected spatial sequences.
The interior of the disc is designed as a dynamic and horizontal working environment, consisting of office spaces of various types and characteristics. A sequence of planes and transparent surfaces is used to enhance interaction between users and intense connections with the external environment. Closed offices with corner meeting rooms and group work niches are located along the exterior facade, opening the central space to the landscape.
Around the atrium, which is the main circulation space, are the entrance, visitor reception area, hot desks, a circular presentation room, and a small auditorium that leads to a multi-functional lounge area and an outdoor terrace in the basement. A spiral staircase connects this area to a small pavilion on the roof, a terrace, and a running track.
The structural system consists of two pre-stressed concrete composite slabs, connected by a series of V-shaped perimeter supports, four reinforced concrete cores, and freely placed groups of inclined supports. Large structural spans allow for flexible spatial organization using light, non-structural, and demountable partitions.
The building creates its own context while interacting with the surrounding nature and the recently constructed neighboring environment. Thanks to the continuous landscape extending beneath the building, the atrium becomes the central point of both internal and external spaces, blurring the boundaries between them.
