Is AI making you stupid?
Over the course of a series of essay-writing sessions, students working with as well as without ChatGPT were hooked up to electroencephalograms (EEGs) to measure their brain activity as they toiled. Across the board, the AI users exhibited markedly lower neural activity in parts of the brain associated with creative functions and attention. Students who wrote with the chatbot's help also found it much harder to provide an accurate quote from the paper that they had just produced.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Australian
5 hours ago
- The Australian
Effective strategies for turning cyber risk data into business insights
Business stakeholders, from board members and C-suite executives to regulators and auditors, seem to be looking for answers as to how they should view cyber risk in the context of their role. 'In today's AI-driven landscape, traditional methods of manually gathering technical cyber data points and attempting to report the same information to various audiences across business, technology, and cyber leaders are no longer effective,' says Ajay Arora, a managing director with Deloitte & Touche LLP. 'Organisations should embrace advanced analytics and tailored communication strategies to confirm that cyber risk insights are meaningful and actionable for each stakeholder group.' 'Board members are asking whether the company is exposed to cyber risks that they're reading about in the newspapers,' says Raj Mehta, a partner with Deloitte & Touche LLP. 'Many CIOs and CFOs are asking whether company investments in cyber security capabilities are aligned with the industry and peers. And regulators and auditors are asking whether the organisation has put the right tools and processes in place.' As the face of cyber security for the organisation, chief information security officers (CISOs) face multiple challenges when it comes to answering these questions, be it the manual nature of data collection, the complexity of generating meaningful analytics or differing opinions and comfort levels when it comes to risk representation. 'Risk decision-making around cyber should be as credible, defendable, and trustworthy as financial statements,' says Ajay Arora, a managing director with Deloitte & Touche LLP. 'Finance teams collect data that is spread across different tools and processes in the organisation, and there is a clear methodology, framework, and understanding of what statements and reports like a balance sheet and a P&L are. That enterprise-wide understanding does not typically exist in cyber today, but it is where the industry is going.' To get to the point where cyber is a standard part of the general business lexicon, it is essential for an organisation to develop a common assessment methodology for the risk-relevant data generated by cyber security tools and processes. 'Risks can then be quantified by consolidating and normalising data for processing through a common risk model,' says Mehdi Houdaigui, a principal with Deloitte & Touche LLP. 'The output of the model is then used for the purpose of creating detailed analyses across business units, regions, and functions in a way that is meaningful to different audiences.' Below are three strategies to consider when building a foundational and trustworthy cyber reporting capability that enhances understanding for stakeholders. Build a scalable cyber analytics foundation The first step is to understand the audience and their use cases. Consider separating stakeholders into broad categories — for example, the cyber team, the IT team, and an extended business category — and then segmenting those categories further into different levels, such as executive, management, and operational. When it comes to building a cyber analytics foundation, the first element to put into place is a metrics framework that incorporates the appropriate key risk indicators (KRIs), key performance indicators (KPIs), and the underlying data points to support each. For example, a reporting program that indicates workforce cyber resilience by identifying trends in failed phishing tests, or data loss event resolution. It might also provide supply chain risk intelligence by focusing on program governance and assessment coverage and remediation. 'To run a cyber metrics and reporting program, organisations will need to continually analyse the data sets collected from their portfolio of cyber and technology tools,' says Stephen Gathman, a manager with Deloitte & Touche LLP. 'From those data sources, an effective set of risk indicators can be produced as a foundation for communicating cyber-induced business risks,' he explains. Next, using standard risk scoring methodologies, data transformation, and advanced techniques (such as AI and machine learning), develop a risk engine that can take cyber data feeds and translate them into indicators of business risk. Once these foundational capabilities are in place, the organisation can work on maturing capabilities in three areas: effective storytelling that is customised to particular audiences, translating technical cyber risk into both business risk and financial terms, and linking the cyber strategy to the business strategy. Confirm trustworthy data quality, models Technology and application teams must parse through vast amounts of cyber data to address risks, which can lead to uncertainty in prioritisation of risk reduction efforts. Common challenges include conflicting or redundant data gathered from multiple data sources, mixed data structures and models leading to issues when merged, varied rating methodologies and scales that can lead to confusing results, and excessive metrics tracking that can produce unclear messaging in reporting. 'It is imperative for cyber analytics teams to build trust in the quality of the risk analytics and metrics being produced from the varied data sources by implementing consistent and transparent models,' says Duncan Molony, head of Cyber Security and Data Analytics at Corebridge Financial. Some trust-related metrics include identity and access protocols that help leaders visualise increased attack vectors, or secure application development (DevSecOps) and tracking the use of secure code repositories. Several steps can help address these challenges and develop a mature risk model: • Deploy a common data model that houses data from multiple sources to maximise utility. • Normalise the common data model to remove redundancy and achieve a centralised warehouse of risk data. • Leverage a common risk scoring methodology to enable risk aggregation over multiple dimensions, such as business units or applications. 'Traditional cyber metric bottom-up reporting and cyber risk quantification (CRQ) in financial terms are increasingly converging to provide a richer context for decision-makers,' says Molony. 'This integration allows organisations to present a more comprehensive view of cyber risks, aligning technical data with financial impacts to enhance strategic decision-making.' The goal is to end up with data that is accurate, complete, consistent, unique, and timely. This can then be aggregated appropriately for a particular audience. 'As cyber risk reporting moves up the chain within an organisation, the scope of relevant risk metrics narrows and more data needs to be aggregated at the appropriate level for the right audience or stakeholder group,' says Tiffany Kleemann, a managing director with Deloitte & Touche LLP. 'For example, those at the operational level may need to see technical data points, whereas those in the C-suite may require information to be aggregated and presented in business terms that reflect business risk, operational resiliency and disruption, or compliance risk. As aggregation increases, it is imperative to strengthen the foundation of data governance and data quality to build and sustain trust in cyber risk reporting,' adds Kleemann. Provide actionable risk intelligence 'In Deloitte's experience, boards and audit committees often ask questions that fit into one of two buckets: risk exposure, or readiness and resiliency,' says Arora. 'They want to know how exposed the company is to cyber risk and then how ready the organisation is to respond should an incident occur.' A leading way to measure and illustrate cyber risk is to build a series of persona-based dashboards and composite, outcome-oriented indicators that can provide actionable insight in a way that is easily digestible. Effective reporting and dashboards gauge levels of cyber risk exposure and resilience, helping the organisation to quantify its cyber posture via frameworks such as the National Institute of Standards and Technology Cybersecurity Framework. In addition to decision intelligence, actionable cyber reporting can be used to translate cyber risks into the business terms typically used to discuss operational disruption, reputational risk, or financial loss. 'The output also provides cyber teams with the insight to break down items by dimensions, such as business units, brands, products, or regions so the information is meaningful for the owners who drive action in the business,' says Arora. 'Being able to slice and dice the metrics by dimension is what helps make the risk intelligence actionable,' says Gathman. By leveraging momentum from strong quantification foundations and data models, the rapid advancement in AI and data collection is slated to enable streamlined identification and potential burndown of cyber risk. By pursuing these capabilities, organisations can enhance investment value in cyber tools and capabilities, while removing ineffective processes and technologies. Isobel Markham, senior writer, Executive Perspectives in The Wall Street Journal, Deloitte Services LP As published by the Deloitte US Chief Financial Officer Program in the June 14 2025 edition of The Risk & Compliance Journal in the WSJ. Disclaimer This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional adviser. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication. About Deloitte Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee ('DTTL'), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. DTTL (also referred to as 'Deloitte Global') does not provide services to clients. In the United States, Deloitte refers to one or more of the US member firms of DTTL, their related entities that operate using the 'Deloitte' name in the United States and their respective affiliates. Certain services may not be available to attest clients under the rules and regulations of public accounting. Please see to learn more about our global network of member firms. Copyright © 2025 Deloitte Development LLC. All rights reserved.
The Australian
9 hours ago
- The Australian
How retailers can prepare for the rise of AI shopping bots
AI has the potential to dramatically reshape the way customers shop, and industry leaders are scrambling to create the killer app that will dominate the space. Is it here yet? Not quite — but we're getting close. In the past six months, a major e-commerce player, two payment giants, and an AI technology company have all announced or launched early-stage AI shopping bots — intelligent assistants that can search for products, compare prices, read reviews, and even complete purchases on behalf of customers. Once widely adopted, AI shopping bots could significantly disrupt traditional retail models. Historically, the retailers who succeeded were those who best curated and managed physical store inventory, or who most effectively linked a customer's online search to their product range. AI agents challenge these fundamentals. They understand customer preferences more deeply, bypass intermediaries, remove price asymmetries, and handle the more tedious administrative aspects of shopping. This impact is especially pronounced in models like marketplaces and loyalty programs, which rely heavily on range and value. Who wants to spend hours navigating clunky marketplaces if they don't have to? How valuable is a loyalty program's discounts and special offers if an AI bot always finds the lowest price — or integrates its own rewards system? However, these agents are not without fragility. A loss of customer trust from breaches in data and privacy could quickly erode the strength of their value proposition. Harsha Maddipatla is Partner, Retail Strategy, at Monitor Deloitte So, what are the implications for retailers? The bottom line is that standing out on value has never mattered more. In a landscape where digital-native companies are leading the way in building AI agents, most traditional retailers may not even get a chance to compete in this space. That makes doubling down on value critical. With an agent choosing from infinite options, price and value become the deciding factors. Pushing products into the market at the most competitive prices ensures they're surfaced to customers more often. This favours discount retailers but could leave mid-tier retailers struggling. And naturally, any price-driven competition is likely to compress margins across the sector. Retailers with scale and influence may also choose to invest in personalisation, search, or agent-driven websites, enhancing the customer experience within their own ecosystems. Partnering with others, strengthening loyalty programs, and offering high-quality products may help retain customers who would otherwise turn to AI shopping bots. There is also a world where an agent-to-agent economy allows a retailer's website agent to interact with customer shopping agents and offer curated, ring-fenced range and more dynamic pricing to remain competitive. A strategic shift in marketing will also be necessary. If AI agents gain traction, new shopping agent promotional channels will emerge (e.g. preferential placement in draft shopping baskets). Depending on the retailer, this may require a re-allocation of marketing dollars from traditional media channels. Physical stores, too, will need to evolve. While they will likely incorporate AI into omni-channel strategies, they must also offer something unique. To compete with the ease of digital agents, brick-and-mortar retail must become more experiential — entertaining, surprising, and rewarding shoppers who visit in person. This shift could also reinvigorate loyalty programs. For these programs, shopping agents are as much of an opportunity as they are a threat, given that the value of loyalty programs often lies in linking in-store and online transactions to create a rich dataset that can enhance agent functionality. However, failing to adapt could mean losing relevance in online shopping. So, what options do loyalty programs have? One is to build their own AI agents. Many programs already have extensive customer data and might be tempted to compete — especially with the right technology and last mile delivery partnerships. But few loyalty programs span multiple brands, which is a key requirement to compete in this space. Single-brand programs lack the reach to rival AI agents that search across dozens of retailers. Another option is for loyalty programs to integrate with whichever AI shopping agents emerge as leaders. These programs offer key differentiators: rewards and data. The insights loyalty programs provide into customer behaviour across both physical and digital touchpoints could greatly enhance AI agents' relevance and utility. In return, loyalty programs would gain access to a new, high-value channel and could even increase online engagement. The rise of AI shopping agents marks a turning point for retail. Customers will no doubt adopt them with enthusiasm once they figure out how much time and money they can save. To stay competitive, retailers can't stick their fingers in their ears: they must double down on value, improve their digital presence, embrace emerging AI agent channels, and explore how loyalty programs can not only adapt to, but enhance, the agent ecosystem. Harsha Maddipatla is Partner, Retail Strategy, at Monitor Deloitte. - Disclaimer This publication contains general information only and Deloitte is not, by means of this publication, rendering accounting, business, financial, investment, legal, tax, or other professional advice or services. This publication is not a substitute for such professional advice or services, nor should it be used as a basis for any decision or action that may affect your business. Before making any decision or taking any action that may affect your business, you should consult a qualified professional adviser. Deloitte shall not be responsible for any loss sustained by any person who relies on this publication. About Deloitte Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee ('DTTL'), its network of member firms, and their related entities. DTTL and each of its member firms are legally separate and independent entities. Please see to learn more. Copyright © 2025 Deloitte Development LLC. All rights reserved. -
Sky News AU
17 hours ago
- Sky News AU
Robot creates AI portrait of King Charles
Sky News contributor Louise Roberts discusses the AI portrait of King Charles created by Ai-Da robot. 'The portrait of Charles is a tech impression, it's not meant to be a real impression,' Ms Roberts told Sky News host Caroline Di Russo. 'What's missing of course is that wonderful twinkle in Charles eye and that sort of rye expression that we know him for and love to see. 'I don't think we're going to be having robots taking over from the royal portraits anytime soon and oh goodness that's a good thing in my book.'
