Latest news with #AWS-hosted
Time of India
10 hours ago
- Business
- Time of India
Revenge code? Ex-employee in Bengaluru crashes grocery app after layoff
A dramatic cyber breach at Bengaluru-based grocery-tech startup KiranaPro has uncovered a bitter truth in the digital economy of the present day—startups can be as susceptible to insider attacks as they are to outsider cyberattacks. What had seemed to be a sophisticated hack proved to be an act of corporate sabotage by an erstwhile employee who had been fired but still had access to important systems. The breach happened in early June 2025, soon after KiranaPro started layoffs due to financial stress. As per the company's leadership, including CEO Deepak Ravindran, the former employee was able to erase parts of the company's backend infrastructure, such as GitHub code repositories, cloud logs, and some AWS-hosted services. Most importantly, it became possible due to the lapse in revoking access credentials once the employee had made a mistake that costed them dearly. Although the extent of the incident was serious, the company has assured that customer information was not breached. Due to internal backups, especially those located locally by other employees, KiranaPro managed to recover most of its system. Internal operations were disrupted briefly, though no core customer-facing services were directly impacted, though. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like เทรดทองCFDsกับโบรกเกอร์ที่เชื่อถือได้| เปิดบัญชีวันนี้ IC Markets สมัคร Undo The company subsequently lodged a police report and launched legal action against the perpetrator. Security vulnerabilities meets financial stress Although the act of sabotage itself was headline news, the circumstances behind it provide a clearer picture of the dangers many startups ignore. KiranaPro was reportedly struggling with the late payment of salaries to current and former staff at the time of the breach. Although the company hasn't attributed the delay to the sabotage, the timing has raised eyebrows about how financial woes can feed internal discontent. The attack also highlights a rising but underappreciated threat across the tech sector—internal users with admin-level privileges and unresolved grudges. Insiders have an advantage over external hackers in that they know the guts of a system, its vulnerabilities, and where to do the most harm. In this instance, no sophisticated malware or phishing was necessary; only a set of credentials and a motive were enough. The initial assumption by the startup that it had been hacked externally introduced a time lag between finding the real cause. Forensic tests were not done before the team arrived at the conclusion that there was no involvement of an outside entity. The breach was completely homegrown. credit: instagram What do we learn from this? KiranaPro's experience is a case study in the consequences that result when HR procedures and cybersecurity measures do not intersect. First, deactivation of credentials at offboarding has to become business-as-usual, particularly for firms dealing in sensitive infrastructure. Second, multi-level authentication and real-time activity tracking by administrative users have to become business as usual. Third, isolated and encrypted regular backups need to be treated as non-negotiable assets rather than optional layers. Finally, there is the human element. Startups need to understand that financial slowness, communication breakdown in layoffs, and insufficient emotional intelligence in employee transitions can all be building blocks of a poisonous culture, one in which digital revenge will indeed be an outcome. KiranaPro might have restored its data, but the actual warning is elsewhere: in an expanding environment where technicality takes precedence over procedural protection, even a single mistake can be the source of a breach not from the outside but from within.
Straits Times
29-05-2025
- Business
- Straits Times
NIE, Amazon Web Services set up AI innovation hub in NTU to boost trainee teachers' tech skills
AWS Singapore country manager Priscilla Chong showing summit attendees on May 25 local enterprises that are using its services. PHOTO: AWS SINGAPORE – Trainee teachers, students and staff at the National Institute of Education (NIE) can get direct access to artificial intelligence (AI) and cloud technology from industry giant Amazon Web Services (AWS). The institute has joined forces with AWS to set up a Technology for Education Centre at its Nanyang Technological University campus. Their three-year memorandum of understanding (MOU) was announced at the May 29 AWS Summit held at Marina Bay Sands, which drew more than 5,000 attendees. NIE hopes the collaboration – the first for the 75-year-old institute with an industry player – will be the start of more such links so that student teachers will be better equipped in AI and cloud skills. NIE director Liu Woon Chia told a briefing that students will be mentored by both faculty and AWS specialists to apply AI to the real world, such as designing tools for students with special needs. The tech centre, which will be housed temporarily in a converted classroom when the new term starts in August, will eventually have its own premises in a new annex building. AWS training programmes will be included in the curriculum, with students earning credits on completion. There are also plans to co-host hackathons, hands-on workshops and student-led forums to discuss ethical AI in education. NIE director Liu Woon Chia (left) and AWS Worldwide Public Sector Singapore country manager Elsie Tan signed the MOU on May 26. PHOTO: AWS Professor Liu said the centre is 'going to be a space where students, faculty or staff can come in and say, 'Hey, I've got this problem in education that I think we should have a better solution. Now, let's discuss and talk to AWS experts.'' More than 1,000 trainees graduate from NIE each year and go on to teach in Singapore schools. It also enrols master and doctoral degree students. AWS also announced at the conference about its AI Spring Singapore programme, which it launched in 2024 to support Singapore's AI blueprint. The US company, which has operated here for 15 years, has pledged investments and support for the public sector, local workforce, enterprises and start-ups, community and research. Its programme activities over the past year include hosting a sandbox environment for Synapxe, Singapore's national HealthTech agency; providing computing and AI resources to Temasek Polytechnic; and working with AI Singapore to create the Asean Large Language Model League competition. The tech firm, which recorded revenue of US$107 billion (S$138 billion) in 2024, is the world's largest hyperscaler, or large-scale cloud service provider, with an almost one-third share of the market. At the summit, Mr Adrien Desbaillets, chief executive of food chain SaladStop!, demonstrated its AWS-hosted AI assistant LuLu, which it plans to launch in September. The conversational voice bot combines data of popular ingredient pairings, menu availability, customer order history and real-time request to personalise a salad bowl recommendation. It also listed the meal's carbohydrate, healthy fat and protein amounts. When a customer says she does not like edamame, the bot will suggest replacing the green bean with tofu, adding that there is no drop in protein value. SaladStop! plans to go beyond its stores to sell its food direct to organisations such as hospitals and fitness centres. Mr Desbaillets said: 'For us, the chatbot experience is incredible. We believe it's could be so much bigger than our 80 outlets today. We can really take things to a whole new level.' Join ST's Telegram channel and get the latest breaking news delivered to you.
Associated Press
31-01-2025
- Business
- Associated Press
Backplain Makes Open Source DeepSeek (R1) Available Alongside 32 Other Models including ChatGPT (GPT-4o)
On a platform designed for secure, safe, and enhanced use of AI through side-by-side model response comparison from a single, simple-to-use user interface. 'DeepSeek's privacy policy states it collects keystroke data, IP addresses, and even tracks actions outside the app. That alone should be enough to make people think twice…' — Lauren Hendry Parsons, Digital Rights expert SAN DIEGO, CA, UNITED STATES, January 31, 2025 / / -- Backplain, a leading SaaS AI Control Platform, already recognized as an Emerging Specialist in Generative AI Engineering by Gartner®, announced today the inclusion of the open-source DeepSeek R1 model. 'As soon as DeepSeek was announced, our users asked for it to be added to Backplain,' said Tim O'Neal, CEO of Backplain. 'But with secure controlled use of models as our primary product tenet, it was clear that we had to only make the open source version of the R1 model available running on our AWS-hosted infrastructure.' Backplain's platform is designed with security and compliance at its core, supporting content anomaly detection and data protection, both key elements of Gartner's Trust, Risk, and Security Management (TRiSM) framework. As a mediation layer between the organization and any AI model, particularly Large Language Models (LLMs), Backplain validates information flows to and from a model against organization policies to help mitigate content generation risks; this protects organizations from potential legal, reputational, or decision-making risks associated with uncontrolled LLM outputs. Backplain also encrypts, obfuscates, and controls data flow to and from LLMs, ensuring data privacy and confidentiality by preventing the exposure of proprietary or sensitive data to third-party environments. 'With DeepSeek's accuracy already in question, the ability to directly compare its responses with that of other models is going to be critical,' said Reed Anderson, Chief Product Officer of Backplain. 'Time and time again, our users tell us that simple side-by-side comparison of responses makes a huge difference to their productive use of Generative AI/LLMs.' Disclaimer: Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved. About Backplain Backplain is an AI platform to independently control AI models, including Large Language Models (LLMs), across an entire organization. Multi-model aggregation of public and private LLMs in a single, simple interface provides the best response, protection from outages and avoids model lock-in. Monitoring, filtering, and reporting provide AI Trust, Risk, and Security Management (AI TRiSM). Prompt assist, multi-response comparison, and Human-in-the-Loop (HITL) sharing builds better questions, identifies hallucinations, and audits content to drive Productivity gains. Learn more by visiting Kevin Hannah Backplain, Inc.
