CybserSecurity News

Tag: ChatGPT

  • OpenAI launches ChatGPT app for the iOS platform

    OpenAI has recently announced the launch of its official ChatGPT app on the iOS platform, hinting at an impending Android version.

    At present, the iOS version of the ChatGPT app is exclusively available in the United States but is expected to roll out in various other regions within the upcoming weeks. The precise timeline for the release of the Android version remains unconfirmed.

    The app facilitates user interaction through text-based conversations or voice inputs transcribed by the mobile microphone. It offers a quick login feature through an Apple ID or Google account. Use of the app is free of charge, but access to the ChatGPT Plus service, which promises a quicker response time and the utilization of the larger GPT-4 language model, necessitates an additional fee.

    Moreover, the iOS version synchronizes with the desktop version, allowing users to continue their queries from the desktop edition seamlessly. This integration is also projected to be available with the Android version.

  • Global server shipments continued to decline in 2023

    Although the recent period has seen a substantial decline in consumer demand for PCs, a ripple effect has even touched the hitherto stable server market, thereby unsettling the entire semiconductor supply chain. Recent statements from TrendForce, supported by the latest data and research, reveal an anticipated continued downtrend in server shipments throughout 2023.

    In light of cloud service providers successively reducing their orders, server OEM manufacturers have downgraded their annual shipment estimates between February and April of this year. Compounded by international circumstances and economic factors, the outlook for server demand appears bleak. TrendForce anticipates global server shipments will shrink to 13.835 million units this year, marking a 2.85% reduction compared to last year.

    The server market performed poorly in the first half of the year. Coupled with end-inventory corrections, the first quarter—traditionally a slow season—saw server shipments fall by 15.9% quarter-on-quarter. The second quarter did not witness the sharp increases of previous years, with a mere 9.23% quarter-on-quarter growth. Besides the de-stocking of the supply chain and a decrease in manufacturers’ orders, environmental concerns also affected server usage. Companies have adapted to cost-control measures, generally extending the service life of servers, indirectly impacting the market’s procurement.

    The much-discussed ChatGPT has spurred shipments of AI servers this year, with cloud service providers, including Microsoft and Google, actively participating. AI server shipments are predicted to grow more than 10% this year. However, the proportion of AI servers is less than 10% of the entire server market, so their impact on overall server sales is minimal.

    TrendForce suggests that a server market reversal depends on the speed of de-stocking. Ideally, this could occur in the second half of this year, although it may not happen until the first half of next year. This situation affects the introduction of new platforms and lowers the purchasing intent for products like DDR5 memory, leaving the possibility of a further decline in server shipments this year.

  • The growth rate of ChatGPT usage traffic has begun to slow down

    In contrast to its initial launch, which garnered over 100 million users in approximately two months, OpenAI’s ChatGPT has begun to experience a deceleration in usage.

    Similarweb’s website analytics data reveals that after ChatGPT went live in November last year, it accumulated around 266 million connections in December, a traffic volume comparable to that of Yahoo News. In January, traffic increased by 131.6%, followed by a decrease to 62.5% in February, and a further contraction to 55.8% in March. By April, the growth rate had diminished to a mere 12.6%, yet it still boasted 1.76 billion usage instances.

    The rapid decline in growth after a few months suggests that the initial surge of traffic for ChatGPT was due to the novelty of the experience and the numerous businesses conducting tests and analyses. The massive traffic growth in the first few months following the launch was followed by a waning interest and, consequently, a more gradual increase in usage.

    Nevertheless, despite the slowdown in traffic growth, the current 1.76 billion usage instances still surpass the traffic of services such as Microsoft Bing, The New York Times, and CNN within the United States. However, when compared to the more frequently used Google search service, ChatGPT’s traffic constitutes a mere 2% of its volume.

    In the wake of ChatGPT’s debut, an increasing number of autoregressive AI services have emerged, including Google’s “Bard,” which will be further detailed at Google I/O 2023, and the autoregressive AI service that AWS claims to be adopting.

  • Navigating AI Ethics: U.S. Government’s Bold Initiative

    The United States government recently announced the allocation of $140 million for the establishment of seven novel artificial intelligence research and development centers, ensuring that AI technologies do not jeopardize public safety. Subsequently, U.S. Vice President Kamala Harris and White House Chief of Staff Jeff Zients met with Google CEO Sundar Pichai, Microsoft CEO Satya Nadella, and OpenAI CEO Sam Altman, among others, urging technology industry leaders to uphold ethical, responsible, and lawful standards to ensure the safety and reliability of AI products.

    Kamala Harris believes that auto-generative AI technologies such as Bard, ChatGPT, and Bing Chat can facilitate national growth while potentially exposing the country to significant challenges, as exemplified by the technology-induced shift in the 2016 U.S. presidential election outcome.

    AI replace job

    During the meeting with technology industry representatives, the U.S. government hoped to emphasize the risks that AI technologies may pose and explore ways to mitigate such risks. Additionally, the government aims to foster safer AI applications through public-private collaboration, without compromising public privacy or exacerbating conflict and distrust.

    Furthermore, Harris revealed during the meeting that should technology industry players fail to fulfill their responsibilities regarding AI or allow these technologies to adversely impact public life, they would face government intervention, implying a subtle warning.

    The White House statement indicated that the U.S. government unveiled its AI Rights Act plan in October of the previous year, intending to guide technology industry players in designing, developing, and deploying AI and other automated systems to protect American citizens’ rights and safety.

    The U.S. government further stated that six leading AI technology companies, including Google, Microsoft, NVIDIA, OpenAI, Stability AI, Anthropic, and Hugging Face, agreed to have their AI systems publicly assessed during DEFCON 31, held from August 10th to 13th, to determine compliance with the principles set forth by the Biden administration.

    The United States Office of Management and Budget plans to release AI technology usage guidelines in the coming months, enabling businesses to deploy AI technologies according to specific regulations and gather more public input before finalizing relevant policies.

  • OpenAI lost $540 million last year due to GPU purchases, electricity, and labor costs

    Acquiring intelligence from The Information website, it has been alleged that OpenAI’s accumulated losses throughout 2022 amounted to $540 million, doubling in comparison to the previous year of 2021.

    Despite ChatGPT swiftly captivating a user base surpassing hundreds of millions and subsequently introducing a premium version, the extensive usage of GPUs required for accelerated computations significantly amplified the underlying costs for OpenAI to maintain its service offerings.

    Aside from the escalating costs of procuring GPUs and electricity, the persistent recruitment and employment of technical talent have also contributed to the expansion of OpenAI’s expenditures. Thus, accepting billions of dollars in investment from Microsoft and substantially integrating with Microsoft’s Azure cloud services platform aimed to mitigate operational losses.

    As articulated by OpenAI CEO Sam Altman, the ambition for the upcoming years is to actualize a $100 billion fundraising initiative, thereby fostering the advancement of a more expansive range of general artificial intelligence technologies.

    OpenAI has secured a new funding round worth $300 million, elevating its market valuation to approximately $27 billion to $29 billion. Participants in this latest round include venture capital firms such as Tiger Global, Sequoia Capital, a16z, Thrive, and K2 Global. The round also encompasses the Founders Fund, established by PayPal co-founder Peter Thiel.

    Prior to this, OpenAI had entered into a long-term investment partnership with Microsoft, which not only involved incorporating ChatGPT technology into Microsoft’s array of products and services but also granted OpenAI access to Microsoft’s Azure cloud platform and other resources.

  • Samsung completely bans employees from using ChatGPT at work due to data breaches

    Samsung previously experienced a sensitive data leak involving an employee using ChatGPT, leading to a directive prohibiting staff from sharing sensitive information with the AI.

    However, Samsung has now expanded the scope of this ban, allowing employees to use ChatGPT on personal devices and outside of work hours, but strictly forbidding any mention of company information, particularly that which pertains to intellectual property.

    The initial leak involved partial product source code, and it remains unclear what prompted this broader restriction. Samsung mandates that employees adhere to security guidelines, warning that failure to comply and resulting in company data leaks could lead to disciplinary action or even termination.

    Additionally, Samsung is currently developing its own artificial intelligence software that includes a conversational robot, designed to assist employees in summarizing reports, writing software, or translating text. This software, however, remains in development and is not yet in use.

    Samsung’s actions underscore the risks associated with AI, as it is neither the first nor the last company to experience data leaks due to ChatGPT. More businesses may follow suit, imposing similar restrictions on their employees’ use of ChatGPT.

  • OpenAI’s ChatGPT Overcomes Italian Privacy Hurdle

    At the end of March, Italy’s privacy regulator imposed a ban on ChatGPT, citing the collection of nonessential information without age categorization, effectively contravening the European Union’s General Data Protection Regulation (GDPR).

    Subsequently, the regulator demanded that OpenAI cease providing services in Italy, resulting in Italian IP addresses being met with unavailability notifications when attempting to access ChatGPT.

    Recently, Italy’s privacy regulator announced the lifting of the ban on ChatGPT, as the platform has preliminarily addressed the issues pointed out by the regulator following necessary modifications. Post-revisions, ChatGPT has been reinstated in Italy, with users now required to provide their age during registration.

    OpenAI’s solutions are straightforward: first, users can choose whether to share their conversation content with OpenAI for subsequent AI model training, a feature available not only to Italian users but also to those in other markets who can actively disable data sharing; second, users must specify their age when registering for ChatGPT. If the user is under 18 years old, an option will appear inquiring if they are between 13 and 17 years old, requiring parental or legal guardian consent for registration.

    Naturally, one might wonder about users below the age of 13; the answer is that those who enter an age below 13 will be prohibited from registering.

    Nonetheless, since the Italian regulator initiated these changes, Italian users are granted additional “privileges,” such as the ability to view which information ChatGPT uses and how that information is collected for AI training purposes.

    The issues encountered by ChatGPT are likely to arise for platforms like Microsoft’s Bing Chat and Google Bard, prompting these companies to proactively improve their services and avert potential bans from the European Union.

  • OpenAI received a new round of financing of $300 million

    TechCrunch reports that OpenAI has recently secured a new round of funding valued at $300 million, elevating its market valuation to approximately $27 to $29 billion.

    Participants in this funding round include prominent venture capital firms such as Tiger Global, Sequoia Capital, Andreessen Horowitz (a16z), Thrive, and K2 Global, as well as the Founders Fund, co-founded by PayPal’s Peter Thiel.

    While the Founders Fund, Thrive, and K2 Global are first-time investors in OpenAI, Tiger Global, Sequoia Capital, and a16z have previously invested in the company, later divesting their stakes. Their renewed investment underscores confidence in OpenAI’s continued technological advancements, particularly as its ChatGPT drives the adoption of generative artificial intelligence applications.

    OpenAI GPT-5

    OpenAI’s recent trajectory has garnered significant industry attention, spurring the development of large-scale natural language models (LLMs) and generative pre-trained transformer (GPT) applications. However, this growth has also incited increasing concerns surrounding the societal implications of AI development, including potential impacts on privacy, misinformation dissemination, and moral dilemmas.

    In light of these emerging apprehensions, a growing number of industry professionals advocate for more proactive government regulation of artificial intelligence. OpenAI, along with other AI developers, has expressed a commitment to fortifying privacy, and ethical safeguards, and ensuring the accuracy of disseminated information in response to these concerns.

  • FTC calls on U.S. government to step in to regulate surge in fraud using artificial intelligence

    The United States Federal Trade Commission (FTC) recently stated that, due to the increasing prevalence of scams involving artificial intelligence technologies such as ChatGPT, it will begin to pursue those who exploit such technologies for fraudulent activities.

    In comparison to previous applications of artificial intelligence, contemporary generative AI technologies are capable of producing realistic images, more vivid prose, and even convincingly deceptive videos, websites, or programs. Consequently, numerous unscrupulous individuals have started to employ these technologies for scams.

    ChatGPT Plus
    Image credit: Future

    During a congressional hearing, FTC Chair Lina Khan, along with Commissioners Rebecca Slaughter and Rebecca Slaughter, explained that AI technologies could potentially be employed for fraudulent purposes, leading to a sharp rise in consumer victimization. They expressed concerns that AI could exacerbate fraudulent behavior and called on the US government to regulate such technologies, lest they cause even greater harm.

    The FTC maintains that technology companies cannot evade responsibility merely by citing the black-box nature of their algorithms. The commission emphasizes that technology companies should also be held accountable for their innovations.

    Previously, the FTC had issued broad public guidance to technology companies, urging them to develop AI technologies responsibly. However, despite the majority of technology companies claiming to adopt responsible and transparent approaches to AI research and development, it is evident that many situations remain inadequately addressed.

  • Canadian privacy regulator announces investigation into OPENAI and ChatGPT

    Following Italy’s privacy regulatory authority’s investigation into OpenAI and ChatGPT, Canada’s privacy regulatory authority has now initiated similar actions. However, in a slightly different scenario, the Canadian authority claims to have received a complaint accusing ChatGPT of collecting, using, and disclosing personal information without consent.

    The Canadian Privacy Commissioner announced an investigation into OpenAI and ChatGPT, though specific details are not being disclosed at this time.

    The Canadian Privacy Commissioner stated that artificial intelligence and its impact on privacy are of utmost importance, and regulators must stay ahead of rapidly advancing technology.

    Canada ChatGPT

    Currently, the Canadian privacy authority has not released any detailed information regarding this matter, and the agency has stated that it will not divulge further details. The primary objective is to publicly report the investigation’s findings after its conclusion.

    OpenAI has not issued a response to this matter. When local Canadian media contacted OpenAI for inquiries, the company referred to a previously published blog post about privacy protection as an indirect response.

    It is worth noting that there have been previous joint complaints from publishers to regulatory authorities, accusing OpenAI of plagiarism, which entails training AI using content from the internet without obtaining prior approval from copyright holders.

    However, it appears that no regulatory authority has formally initiated an investigation into OpenAI’s content scraping at this time, and the primary focus remains on ChatGPT’s protection of people’s privacy and security.