Opinion: need for AI privacy protections
Everything I touch now has AI embedded in it. The phone, the laptop, every online retailer. And if you listen hard, at least in Dunedin, you can hear the complaints of teachers and lecturers about the AI errors they're seeing in student assignments.
Many Otago education providers are now making money teaching AI, so no irony! And there's to be a $50 million data centre built in Southland which will no doubt host AI processing, so some jobs at least will be created.
Recently, one of the ''inventors'' of AI, Sam Altman, said his children will never be as smart as AI, and let me say: wow, is that worrying!
The AI most of us access is built on the theft of hundreds of thousands of authors' and artists' work, including quite a few Dunedinites, and still produces pictures with six fingers and three hands, struggles with any skin colour darker than mine, and (if it's a large language model) can't add basic numbers.
Among the variations on AI there are now transcription services spreading into mental health and some social services provision; taking notes and summarising conversations between client and worker to reduce the therapist/doctor/social worker's administration time, which is always under extreme pressure.
So, potentially a useful productivity tool to free up time to work with patients/clients. But also, potentially a bit of a privacy risk, and perhaps at risk of some error.
Good questions to ask, if you have a practitioner wanting to use one of these things include:
Where are the recordings and analysis held (on or offshore) and for how long?
What protections does the data storage have and who has access and for what?
Can you have a copy of the transcript and any analysis?
What is the practitioner's policy if you want to correct the transcript and request a review of the analysis?
And then you'll know. Even if the AI doesn't.
• If you would like to support the Mission's work with a donation, please visit www.givealittle.co.nz and search for The Methodist Mission.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Techday NZ
5 hours ago
- Techday NZ
Hitachi Vantara named leader for AI storage in GigaOm Radar 2025
Hitachi Vantara has been recognised as a Leader and Fast Mover in the 2025 GigaOm Radar for High-Performance Storage Optimised for AI Workloads. The newly released GigaOm Radar report evaluates advanced storage platforms designed specifically for artificial intelligence (AI) workloads. This is the first time the report has assessed high-performance storage optimised for AI, and Hitachi Vantara's positioning underscores the company's capabilities in supporting enterprise-scale data requirements for AI and machine learning (ML). Report evaluation The GigaOm Radar for High-Performance Storage Optimised for AI Workloads reviews market solutions based on their key capabilities, foundational criteria, feature completeness, and readiness for future demands. Leaders in this assessment are noted for their maturity and strategic alignment with evolving enterprise AI needs, as well as their ability to execute at scale. Hitachi Vantara's placement as a Leader and Fast Mover was attributed to its Hitachi iQ portfolio, which provides AI-ready infrastructure solutions. The report highlighted several strengths, including quality of service (QoS) and workload isolation, GPU-direct storage integration and AI-optimised data layout and management. The report noted, "Hitachi iQ offers world-class QoS and workload isolation capabilities." By combining file system-level policies with flexible cluster architecture, Hitachi iQ enables effective resource allocation, supporting consistent and reliable performance for high-priority AI and ML workloads, even in multi-tenant or shared environments. The platform's GPU-direct storage integration was identified as another area of strong performance. Optimised drivers within Hitachi iQ support efficient data transfer between storage and GPU memory, enhancing outcomes for AI and ML workflows. Additionally, Hitachi iQ's data management strategy utilises intelligent placement across storage tiers, such as TLC/QLC NVMe flash storage for high performance and object storage for additional capacity. Real-time monitoring and user-defined policies are built in to balance performance and cost efficiency for varying data patterns. Industry commentary "What really stood out about Hitachi Vantara's offerings is the quality of service and the ability to isolate workloads," said Whit Walters, Field CTO and Analyst at GigaOm. "They've delivered a well-integrated, scalable platform in Hitachi iQ, backed by enterprise-proven storage. This combination gives organisations a powerful and flexible foundation for operationalising AI at scale, especially as large-scale AI and GenAI workloads will require the ability to manage data and performance as demands continue to grow." The GigaOm report also referenced ongoing collaborations including a partnership with NVIDIA on the NVIDIA Data Platform for Hitachi iQ, a partnership that adds to the capabilities of the platform. The roadmap for Hitachi iQ includes new hardware and AI solutions, among them the Hitachi iQ M Series, announced earlier in the year. Integration with the Virtual Storage Platform One (VSP One) enables further intelligent tiering between NVMe flash and object storage, providing additional flexibility and performance optimisation. Octavian Tanase, Chief Product Officer at Hitachi Vantara, commented on the recognition. "AI isn't just pushing the boundaries of what infrastructure needs to do; it's completely redrawing them," he said. "Our goal with Hitachi iQ is to give customers a high-performance foundation that removes complexity, accelerates outcomes, and adapts to whatever their AI journey requires next. By integrating Hitachi iQ with our VSP One platform, we're enabling a flexible, intelligent storage strategy that's ready for what's now and what's next." Ongoing awards This rating follows recent industry recognition for Hitachi Vantara. In May 2025, the company was awarded the Sustainable Technology Award at the Global Tech & AI Awards for its efforts in sustainable data infrastructure with the VSP One Block solution. Earlier in the year, GigaOm also recognised Hitachi Vantara as a Leader and Outperformer in the GigaOm Radar for Primary Storage, related to its VSP One hybrid cloud data platform. The GigaOm Radar for High-Performance Storage Optimised for AI Workloads aims to shed light on platforms capable of addressing the increased performance, scalability, and operational requirements integral to enterprise AI and machine learning deployments.
Techday NZ
5 hours ago
- Techday NZ
Businesses embrace AI but lack training & infrastructure for growth
A recent global survey by TeamViewer has found that while small and mid-sized businesses (SMBs) are adopting artificial intelligence (AI) at a rapid pace, their overall maturity in using the technology remains limited, with a particular need for better training and infrastructure. The survey drew responses from 1,400 business leaders worldwide, including 200 based in Australia, and highlighted both the promise and persistent challenges of AI adoption for smaller organisations. The majority of SMB decision makers—95%—acknowledge they need additional training to use AI effectively, despite 72% describing themselves as AI experts. Productivity and insight Australian business leaders reported notable benefits: 33% said increased efficiency and productivity were the main advantages, and a further 28% cited improved insights into processes and performance. However, challenges remain substantial. Respondents identified the lack of education on using AI, perceived security or legal risks, and high implementation costs as their three key concerns. The Australian findings also showed that 70% of business leaders see AI as critical to business innovation and growth, with 67% specifically concerned about the security risks related to data management and 74% stating they use AI cautiously with security measures in place. Additionally, 59% believe AI will have a positive impact on their revenue growth in the coming year provided it can be scaled effectively, while 68% expect AI will drive the most significant productivity increase in a century. Adoption is common, but depth lags The survey found that 86% of SMB leaders globally are comfortable with AI tools being used by employees outside of IT. Yet, regular use remains limited; only one third of SMB respondents use AI daily, and just 16% do so weekly. Despite this, 35% of SMB leaders describe their use of AI as "very mature", compared to just 22% among larger enterprises. Failure to embrace automation through AI is viewed as a tangible risk. For 28% of SMB respondents, increased operational costs resulting from missed automation opportunities are the biggest consequence of inaction. In comparison, larger loss of competitive edge is a greater concern elsewhere, cited by 26% of the wider business community. Security and training Despite optimism about AI's potential—72% of SMB leaders globally expect AI to generate the greatest productivity boom of the century and 76% say it is essential to business performance—skills and security gaps persist. More than one third (38%) say insufficient AI training is a significant hurdle, with 74% expressing concern about data management risks and 65% only using AI in tightly controlled security environments. Furthermore, 77% would not risk a week's salary on their business's ability to manage risks such as unauthorised AI tool usage. Infrastructure and investment Infrastructure readiness is also a major issue. Nearly half (47%) of SMB decision makers say their current systems are insufficient to scale AI as swiftly as they would like. Still, signs of further investment are apparent: 75% of SMBs plan to increase their AI investment in the next 12 months, with the same proportion expecting to ramp up spending within six to twelve months. This suggests a move from initial experimentation towards more advanced implementations. Focus on practical tools TeamViewer has introduced TeamViewer Intelligence, featuring session insights and analytics, alongside its digital assistant, TeamViewer CoPilot, which is designed to help IT teams improve efficiency during support sessions. This tool aims to enable agents to stay focused, automate tasks, and receive real-time guidance without switching applications. "SMBs are clearly motivated to embrace AI, but many are still searching for the right way to turn early adoption into lasting impact," said Artus Rupalla, Director of Product Management at TeamViewer. "The key isn't just more tools, but smarter integration – solutions that bring automation, insight, and consistency into everyday operations. This research confirms what we're seeing across our customer base: SMBs want AI that solves real problems, not just theoretical ones. With practical tools like TeamViewer Intelligence, we can help these businesses move from experimentation to execution and drive real performance gains." The TeamViewer survey report focuses on respondents from businesses with 200–999 employees, and its findings underline the continued challenges faced by SMBs striving to keep pace with AI developments while addressing skills and infrastructure limitations.
Otago Daily Times
13 hours ago
- Otago Daily Times
Opinion: need for AI privacy protections
Everything I touch now has AI embedded in it. The phone, the laptop, every online retailer. And if you listen hard, at least in Dunedin, you can hear the complaints of teachers and lecturers about the AI errors they're seeing in student assignments. Many Otago education providers are now making money teaching AI, so no irony! And there's to be a $50 million data centre built in Southland which will no doubt host AI processing, so some jobs at least will be created. Recently, one of the ''inventors'' of AI, Sam Altman, said his children will never be as smart as AI, and let me say: wow, is that worrying! The AI most of us access is built on the theft of hundreds of thousands of authors' and artists' work, including quite a few Dunedinites, and still produces pictures with six fingers and three hands, struggles with any skin colour darker than mine, and (if it's a large language model) can't add basic numbers. Among the variations on AI there are now transcription services spreading into mental health and some social services provision; taking notes and summarising conversations between client and worker to reduce the therapist/doctor/social worker's administration time, which is always under extreme pressure. So, potentially a useful productivity tool to free up time to work with patients/clients. But also, potentially a bit of a privacy risk, and perhaps at risk of some error. Good questions to ask, if you have a practitioner wanting to use one of these things include: Where are the recordings and analysis held (on or offshore) and for how long? What protections does the data storage have and who has access and for what? Can you have a copy of the transcript and any analysis? What is the practitioner's policy if you want to correct the transcript and request a review of the analysis? And then you'll know. Even if the AI doesn't. • If you would like to support the Mission's work with a donation, please visit and search for The Methodist Mission.
