Magazine menu Close Magazine menu

Artificial intelligence

Latest stories

interconnected technologies ai, occupation, professionals, people wearing helmets, industrial

The next big business challenge: How organizations are preparing to secure AI

  • min read
    • Of all investments today, artificial intelligence (AI) is transforming enterprises, enabling them to provide improved services, automate processes and make better decisions. Business leaders must invest in the right tech at the right time, and have the people and systems in place that can guarantee success. AI will change most industries but also brings new risks and challenges in securing enterprises and safeguarding customers. Kaspersky's new research aims to help businesses stay ahead of the changes AI brings, published in this 2024 report, Connecting the future of business: How leaders should prepare for using and securing AI and interconnected technologies. About the interconnected tech research Kaspersky surveyed 560 IT security leaders in six regions – North America (US,) Latin America (LATAM: Brazil, Chile, Columbia and Mexico,) Europe (Austria, France, Germany and Switzerland,) Middle East and Africa (META: Saudi Arabia, South Africa, Turkey and United Arab Emirates,) Russia and Asia-Pacific (APAC: China, India and Indonesia.) Respondents were either aware of or involved in cybersecurity decisions about interconnected tech such as AI, web 3.0, data spaces, digital twins, augmented reality (AR,) virtual reality (VR,) 6G and internet of things (IoT.) They worked in organizations with at least 1,000 employees across many sectors. Kaspersky asked how they're introducing and securing AI and other interconnected technologies. Here's what the research revealed. AI is organizations' significant focus for the next two years AI is becoming a cornerstone for organizations striving to stay ahead thanks to its ability to foster innovation. In cybersecurity, AI offers predictive analytics and greater adaptability. Goldman Sachs says global investment in AI will reach 200 billion US dollars by 2025. AI is more than a technology. It's a strategy that lets organizations navigate today's complex business environment. AI adoption Nearly all organizations surveyed have already adopted (54 percent) or plan to adopt AI within two years (33 percent.) Respondents see Generative AI (GenAI) as valuable for improving security technologies' performance. Goldman Sachs economists Joseph Briggs and Devesh Kodnani say GenAI could boost global labor productivity by more than one percentage point yearly once its use becomes widespread. IT companies worldwide are upskilling their workforce and undertaking GenAI pilots to avoid falling behind. IT giants Accenture pledged three billion US dollars to data and AI investment, while French tech firm Capgemini is putting two billion Euros towards GenAI over the next three years. Securing AI 49 percent of organizations said they were extremely well, or well prepared, to secure AI. Almost 50 percent think AI is slightly or not at all hard to secure. Compared with securing other interconnected technologies, most organizations felt more prepared to secure AI. In responding to Kaspersky's survey, the head of cybersecurity at a Brazilian bank said, "I mostly don't find security problems with GenAI. We already understand how to run and train it." While many organizations feel confident in their ability to secure AI, a little over half felt less than well-prepared to secure AI. These businesses need to do more to prepare for the inevitable scale-up of AI-based technologies. AI makes digital trust and data compliance indispensable AI introduces new challenges and complexities in protecting sensitive data and maintaining digital trust. Digital trust is an aspect of customer trust. In Kaspersky podcast Insight Story, emerging tech security expert Malek Ben Salem described digital trust as, "Our confidence that our privacy is protected and secured, the reliability of digital transactions and how sure we are of the identities of those we engage with." Tech researchers McKinsey's 2022 Digital Trust report said 70 percent of people trust companies to protect their privacy, but most companies aren't meeting their expectations. Data breaches, unauthorized access and privacy violations can erode digital trust and lead to reputation damage, lost business and regulatory consequences. Building digital trust takes time but is essential to users' confidence in AI. How AI impacts digital trust AI lets organizations collect and store more data about customers, operations and interactions. This wealth of data helps gain insight but also means more risk of breaches. AI facilitates sharing data between devices, networks and organizations. Data sharing aids collaboration, efficiency and innovation but also means more opportunities for unauthorized access. As it is relatively new, organizations may struggle to be clear about their practices and security around AI. Without clear communication, trust may break down. Top ways organizations are improving digital trust and compliance Adopting zero-trust frameworks is organizations' top response to the need to improve digital trust, with 81 percent already using or planning to adopt zero-trust within two years. Other actions include pursuing data privacy and cyber resilience regulations (74 percent) and digital sovereignty initiatives (73 percent.) Kaspersky's research found companies introducing zero-trust frameworks are better prepared to secure AI and other interconnected technologies: 51 percent of zero-trust framework early adopters said they were extremely well-prepared or well-prepared to secure these technologies, compared with just 27 percent of the rest. The zero-trust canvas: Painting the future of cybersecurity The rise of AI is opening the door for organizations to break with all traditions that don't work for today's tech realities. Traditional cybersecurity is vulnerable to insider attack because it relies on trusting those within a network's perimeter while viewing outsiders as suspicious. Enter: Zero-trust frameworks. The head of IT at cloud provider Searce said, "The zero-trust security model is the only framework that works. The traditional security model can't do the job, especially for interconnected technologies." Under zero-trust frameworks, every user and device must undergo continuous multi-factor authentication (MFA) and authorization, regardless of their trust level. Users and devices get the minimum access needed for their tasks, with constant monitoring of user activity, device behavior and network traffic. Regulation and compliance With each territory having different data protection and privacy legislation, global organizations find it hard to achieve consistent standards around AI. 62 percent of leaders find compliance certification hard. They want to see standardized compliance processes across regulations and say online tools and platforms, with clear guidelines, would further simplify certification. Handling growing challenges and preparing to secure AI With the scale of change AI will likely bring, organizations must have a strategy to prepare themselves to adopt and secure it. Kaspersky research uncovered four effective strategies to ensure readiness to secure AI and other interconnected technologies: Adopting security by design principles, training and upskilling your workforce, upgrading your cybersecurity solutions and meeting regulation and standards. The full report, Connecting the future of business, details each strategy. Interconnected technologies like AI bring huge business opportunities but also a new era of vulnerability to serious cyberthreats. With more data collected and transmitted, cybersecurity measures must get stronger. Organizations can use effective strategies to safeguard critical assets and fortify customer trust amid a growing interconnected landscape. Leaders must also resource cybersecurity well enough to enable access to improved cybersecurity solutions that can meet oncoming challenges.
insight story, ethical and ai, robot communicates with people

As AI’s influence rapidly expands, here’s how business ethics must keep up

  • min read
    • We live in a world where algorithms can make decisions and data fuels innovation. It means ethical considerations are more critical than ever for business. They must balance using new technology to increase competitive advantage while preserving integrity and protecting customers. In our podcast Insight Story, experts Tomoko Yokoi (Switzerland,) senior business executive and researcher at Global Centre for Digital Business Transformation, IMD Business School and Andy Crouch (UK,) consultant and co-founder of ethical-AI natural language processing company, Akumen, outline how AI biases can impact business and what steps they can take to ensure its fairness. Kaspersky Global Research and Analysis Team's Dr. Amin Hasbini expands on the privacy and responsible data use implications. Not all AI is created ethically equal Andy's company Akumen found a problem that needed solving. "Scores out of five are useful, but we wanted insight from written responses like product reviews, and there was no way to do it. The team created an AI solution to identify meaning like topics, emotions and sentiment. Sentiment measures opinion – positive, negative or neutral – but emotions drive behavior. It works on text feedback anywhere, which might be about consumer goods, healthcare or anything else." Their approach uses AI differently from generative AI tools like ChatGPT. "Our AI is rule-based, human-created and human-curated. It's completely transparent and there are no algorithms as with large language models. We can dive in and make rules more nuanced if we recognize bias. With large language models, that would be complex and expensive." Andy expands on generative AI's limits for truly understanding people. "We asked ChatGPT how many emotions humans experience – it said 138,000. That doesn't help us understand what drives behavior. Our platform has 22 emotions – enough to see what drives behavior. Through our partner, Civicom, we're helping the UK's national health service (NHS) to understand what patients and staff experience." And that understanding can improve lives: Larger language models use big data pools, but there are also more contained, enterprise-level tools like ChatGPT Enterprise that businesses can furnish with their own data and control how they use it. Tomoko sees enterprise-level tools as useful but notes they can't do what big data can do. "Organizations are developing new functions around AI, like data annotators, who clean data before it goes into models. But is it foolproof? The beauty of using data from everywhere is it gives you insights you otherwise wouldn't get." Choosing ethical suppliers Luckily for companies using AI ethically, more businesses are adopting digital responsibility policies and choosing ethics-first suppliers. Tomoko gives an example. "Deutsche Telekom has been a pioneer in AI ethics. They've trained all employees to ensure AI ethics are distributed throughout the organization. At the same time, they have about 300 suppliers and ensure it's in all their contracts. So it goes beyond the boundaries of the company." But many businesses don't know where to start. Tomoko says, "Over 250 companies have committed to AI ethics, but codified mechanisms only help if they change behavior. How can we live these principles and ideals? External experts can help, and there's a case for individuals taking responsibility, which will have a collective impact." She suggests how companies frame AI ethics matters. "You can see AI ethics as value or as compliance. If it's compliance, it will be cost- or risk-driven. But AI ethics could also be a competitive advantage." Andy compares AI ethics to health and safety. "If you have a health and safety director, it's only one person's responsibility. Change won't happen unless everyone understands health and safety's importance, and especially that it drives productivity and revenue." The competitive advantage is real. McKinsey research found 72 percent of customers considered a business's AI policy before making an AI-related purchase. Tomoko highlights the importance of backing up policies with action. Which AI issues should companies care about? Tomoko outlines three places to look. "First, consider the software development lifecycle. If you're considering developing an AI product, think of how it's designed. Look for bias in the data. "Second, once it's being developed, although many companies say they're implementing AI ethics, people developing AI-driven products don't know how to apply those principles. So, look at how people use ethical principles in day-to-day software development. "Third, we test products in controlled environments. Once it launches, ask who is monitoring it and how we ensure it doesn't gather bias and that people use it correctly." Tomoko is part of IMD Business School and knows that what future executives learn about AI ethics will shape future companies' ethical behaviors with AI. She says, "First, we say everyone has a responsibility to these issues that goes beyond the company. You need to be aware of this responsibility, but also be able to make others in your team aware." Secondly, "What type of organizations do we want to build? We coach people to be able to handle multiple goals – not only profit but also social, environmental and ethical goals. We want them to walk away thinking of the future." Andy drills down into the data AI is using. "Understand how the AI model is built. Is the data you're analyzing through that AI model ethically sourced, and are you using it ethically? The lack of transparency over large language models is rife for ethical risk and bias." AI training data bias can have life-threatening impacts. Poor AI translations have been found to be jeopardizing asylum claims. Andy sees retrieval-augmented generation (RAG,) which uses more proofed datasets, as part of the solution. Can we have secure and well-regulated AI? Dr. Amin Hasbini. Head of Research Centre, Middle East, Turkey and Africa for Kaspersky Global Research and Analysis Team, thinks AI ethical standards are needed. "AI won't self-define its ethics. They must be programmed with ethical standards." Since there is almost no way the public can evaluate, critique or improve AI ethics, regulation must play a part, according to Amin. "We need security and safety by design, and continuous verification of it. That would require transparency, especially from big tech vendors, and letting the public influence how these technologies develop." He likens the challenge to that of regulating social media. "We're asking people to adopt technologies that can do much damage without giving them ways to ensure that doesn't happen. The same has happened before with social media, with it being used for data leaks and fake news. European Union regulation is moving fast around AI, but AI could be much more dangerous than social media – we need rules now." For improved ethical data use, Amin recommends asset management controls. "If well deployed, asset management controls allow data to be classified, including which is available to AI, which can be shared publicly and which needs to stay inside the organization." Andy says regulation is hard in this fast-moving space because no one knows what's coming next. "I question anyone saying they know what will happen in the next six months or beyond. But there's a lot of fear and lobbying going on – so go slow. If your AI-driven capability can't deliver because it's non-compliant, ethically or otherwise, it will be damaging." However, he believes regulation is necessary. "It will be interesting to see how they regulate something that's not easily defined and morphs quickly, but we must protect those who need protecting." Kaspersky has recently proposed six principles for ethical use of AI in the cybersecurity industry with transparency at the core. Getting started with AI ethics Our experts have straightforward advice for those business executives yet to approach AI ethics. Tomoko says, "As a mindset, remember the analog and digital worlds are the same. Your analog-world values should extend into the digital world." Andy highlights the need for both widespread knowledge and deep expertise. "Get your whole team conversant with AI, but have a well-informed friend who lives and breathes this stuff to call when there are challenges." With headlines about AI taking our jobs and AI founders like Geoffrey Hinton sounding the alarm on unregulated AI perils, it's easy to write off AI ethics as a problem too hard to fix. But these complex issues need priority. There are green shoots of change. In December 2023, the AI Alliance launched to focus on developing AI responsibly, including safety and security tools. Its 50 members include Meta, IBM, CERN and Cornell. The message may be, 'Let's not move too fast and not break things.' With OpenAI, creators of ChatGPT, not invited to the party, could the tortoise of collective corporations beat the nimble hare of innovation? AI gives business potential for great gains but comes with great risks to reputation, security and privacy. With strong ethical AI policies translated into action and widespread knowledge among employees, businesses can have more confidence to take advantage of AI's many benefits.
insight story, employer uses generative ai

How businesses can use generative AI to empower employees

  • min read
    • Headlines about the power of generative artificial intelligence (gen AI) are everywhere. But business leaders are asking, is it really that game-changing, or just a fad? And if it is a game-changer, how can you make the most of it? In the first episode of our second season of Kaspersky's podcast Insight Story, I speak with experts Shagun Sachdeva (India,) Project Manager for Disruptive Tech at business intelligence service GlobalData Plc and Karen Quinn (UK,) Senior Director, Brand and Corporate Communications at financial software providers, Finastra. What is generative AI? TechTarget defines gen AI as any "artificial intelligence technology that can produce various types of content, including text, imagery, audio and synthetic data." Some applications rely on Natural Language Processing so users can ask for outputs using everyday speech. Well-known examples of gen AI include OpenAI's chatbot ChatGPT and Google Bard, but current and potential business uses are near endless. Shagun thinks many businesses already understand its importance. "A recent GlobalData Plc survey found more than 50 percent of businesses expect gen AI will tangibly disrupt their industry in the next five years. 30 percent are already using gen AI tools." How are businesses using gen AI? "Business leaders are having more sophisticated conversations about how gen AI can go from an experiment to giving strong returns on investment," says Shagun. "Whether it's chatbots engaging with customers, sharing financial tips and customizing financial plans based on individual spending habits. It's also generating loan options based on a user's credit profile." Finastra plans to develop gen AI tools for customers in future, but they've started in-house with GENAI (X,) a gen AI-based system now rolled out to their 8,000 employees. "We use it for things like understanding and reviewing contracts, predicting behaviors and optimizing workflows," says Karen. She believes that although only being used internally so far, their gen AI tools are benefiting customers. "We can only work with the data we have to make financial decisions. Anything that can generate new scenarios, pull in more data sources or enable federated learning can only benefit." Is gen AI a threat to jobs? Shagun acknowledges employees worry AI might replace them. "Some of that anxiety may be justified – a Goldman Sachs report in March 2023 said AI could replace 300 million full-time jobs." But getting to know gen AI better may help overcome that concern. "Lean in to the technology. Education and training is key. Start with courses like prompt engineering." Karen refers to a quote from economist and professor Richard Baldwin, "AI won't take your job. It's somebody using AI that will take your job." In other words, it's not the technology itself that's a threat to jobs but failing to explore it and take full advantage of it. She advises looking at the benefits of gen AI broadly. "This is not an efficiency play. People will become more productive, but hopefully that means work is more rewarding." Getting gen AI right While generative AI offers much possibility, it also raises concerns around ethics, bias and data privacy. It is largely unregulated today and developing AI models can be resource-intensive. Karen highlights Finastra's caution. "We rigorously tested the products, then launched a full-scale training program and disabled other tools. They're still going through testing, like panels to try and break them and make sure no sensitive data gets shared with the wrong audiences." She continues, "In brand and communications, there are copyright issues. If you're using text-to-image prompts, where is it drawing data from? We need to be very careful." Materials used to train AI are often copyrighted, and whether people or businesses can copyright outputs of gen AI is subject to legal challenges in many countries, with some courts – including in the US and Italy – ruling no, at least in the case of visual arts. Balancing gen AI optimism with realism Shagun feels misinformation is the biggest business risk from gen AI. "It can fabricate 'facts' – what Google researchers call hallucinations." She also recommends businesses pay attention to copyright. "Disputes have arisen between artists and AI companies over the value of human creativity." Data privacy and security also rate highly for Shagun. "Large language models are trained on vast swathes of internet data. There's no data protection embedded in these systems by design or default. Training data can fail to include women, older people or marginalized groups – that's an ethical challenge." Security and generative AI Amin Hasbini is Kaspersky's Head of Research Center for its Global Research and Analysis Team (GReAT,) Middle East, Turkey and Africa. He says many employees are already using freely available AI tools, potentially without their employers knowing. "In a 2023 Kaspersky study, we found 57 percent of workers are using generative AI to save time. This raises many security questions. What kind of data are they putting into it? Is it intellectual property, code or documents to summarize? That's a major concern." Cybercriminals are also using generative AI to help them fool people with more realistic fake websites. In 2022, Kaspersky's anti-phishing system blocked more than 500 million attempts to access fraudulent websites – a doubling of attempts in the previous year. It's likely cybercriminals are already using generative AI chatbot ChatGPT's ability to produce convincing texts to create automated spearphishing attacks – phishing targeted at specific people. To protect against generative AI-based threats, Amin advises, "It starts with awareness. Employees need to know their organization's boundaries around putting data into AI websites. Businesses need enterprise cybersecurity solutions that allow monitoring and control of devices, systems and data used within the organization." Leading employee cybersecurity awareness approaches include implementing a cyber champions program or targeted education as used by Heathrow Airport. Gen AI's future potential What kinds of applications for gen AI will we see businesses pursuing in the near future? Shagun believes what businesses do today will decide the future of gen AI. "The world will look completely different by 2030. Global Data Plc estimates the AI market will be worth around 900 billion by 2030. But that doesn't mean carelessly diving in. "Gen AI is no magic bullet. While the enthusiasm around it is justified, prudence is imperative. We need responsible innovation," says Shagun. Karen imagines gen AI helping to make markets fairer. "Gen AI has incredible potential to overcome inefficiencies, like small-to-medium business (SMB) access to trade finance. Gen AI might be able to generate or automate some of the processes for onboarding smaller companies into global trade, opening them up to a wider audience." She also thinks gen AI could help solve some of the problems it's been known to introduce. "It can help identify bias through explainability and ensure we close feedback loops that cause biases. It could also look at efficiencies in supply chains, contracts and communications and marketing. It can take out some of the drudgery and enrich the processes." "We're on the cusp of understanding what gen AI can do," Karen says. "Who knows what the future brings? This could be game-changing in ways we can't even imagine." Like Karen and Shagun, as a creative marketer, I'm drawn between mostly optimism and a little fear about what gen AI will bring. As I wrote following a gen AI talk for marketers where I heard about Karen's exceptional AI program at Finastra, I'm optimistic we can do it right. Having the training, technologies and policies to bring the whole organization on the journey, we can use this tech to free us by shifting mundane tasks down to these somewhat smart bots. But in that shift, we may narrow opportunities – particularly for talent entering our industries – to refine their craft and learn the difference between average and distinctive output. Refined talent that generative AI lacks. For now, anyway.
chatGPT cybersecurity regulation

With ChatGPT’s vast innovative potential, regulators must take care

  • min read
    • ChatGPT and other artificial intelligence (AI) chatbot applications are major technological milestones with far-reaching benefits for many industries. But with each advance comes unintended consequences. With ChatGPT in its fourth iteration, and its underlying technologies growing fast, it's time to take stock of the policy landscape to understand its implications. What is ChatGPT, and how are businesses using it? ChatGPT is an AI chatbot made by OpenAI. It uses natural language processing (NLP) to create human-like text responses. Users can ask ChatGPT almost anything and receive relatively natural-sounding responses. It can write outputs like programming code, social media posts, summarize complex topics or almost anything text. ChatGPT is an example of generative AI: Artificial intelligence that learns to generate original content like images, music and text by identifying underlying patterns in existing examples. "Its primary objective is to create interactive and realistic conversations with users, making it a powerful tool for chatbots, virtual assistants and customer support applications," says AWS Solutions Architect Ragu Kuppannan. Many top companies are using ChatGPT. For example, travel outlet Expedia wants customers to be able to book a holiday in a way that feels like having a chat with a travel agent. Microsoft, a major investor in OpenAI, says ChatGPT already powers its search engine, Bing, and plans to add it to Word and Excel. ChatGPT could be promising for any application where natural, conversational interaction between people and tech would be beneficial. Risks of ChatGPT Critiques of ChatGPT and other chatbot applications generally focus on its potential to disrupt human employment, produce incorrect information or poor practice in training algorithms to identify harmful content. The underlying cybersecurity risks get less attention. Kaspersky's specialist threat intelligence teams have identified three main risk areas. First, ChatGPT can provide advice, even detailed guides, to help prospective attackers plan and target victims. Second, it can develop code and even create entire coding projects for those with little knowledge of software development. While ChatGPT has restrictions and safeguards to avoid abuse, they're fairly easily bypassed. Third, ChatGPT produces convincing and accurate text that can be used in phishing or spearphishing. To some extent, these three areas represent theoretical risks or outlier scenarios that don't represent how people and businesses usually use ChatGPT. Policy perspective beyond risks Policy should aim to be technology-blind – taking a similar approach to all related technology that might emerge in the future, rather than creating a unique approach for one of today's concerns. It should build the best possible defenses against the worst human intentions and applications. Practically speaking, there may be no way to prevent or limit use of tech like ChatGPT. And to limit an emerging technology's potential means being left behind in a global digital economy. While policy discussions usually focus on outputs – the material ChatGPT produces – its inputs (the data it's trained with) may be a bigger threat. Data-driven tools like ChatGPT are only as good or accurate as their training datasets. As chatbot systems may become the most common way to use the internet, issues will have wide impact. Ensuring balance in training data is a growing challenge as disinformation expands. Bad actors can influence what users see and read in ways that have global consequences. Datasets, devices and systems must be secure and accurate. How nations are regulating AI In our public policy work, Kaspersky advocates public-private partnerships. We think partnerships combining global private-sector expertise with policymakers' local needs will be the best defense. It's a middle-ground approach that avoids too much or too little caution. These partnerships in cybersecurity maximize the advantages of emerging technologies like ChatGPT, while safeguarding against risks. United Arab Emirates (UAE) is a role model for inclusive public-private partnerships. They've taken a progressive and collaborative stance on partnerships with international tech firms. Initiatives and forums like the Dubai Chamber of Digital Economy and the UAE Cybersecurity Council aim to safely and securely advance technological growth for short- and long-term benefit. In May 2023, Brazil was considering a Bill to regulate AI. It requires "systems [to] undergo a preliminary assessment carried out by the suppliers themselves, to determine whether they can be classified as being of high or excessive risk," according to OneTrust Data Guidance. It also restricts exploitative AI uses, like subliminal techniques that could cause harmful behavior or target vulnerabilities in older and disabled people. The Center for Strategic and International Studies says Japan looked at the European Union's sterner approach to AI regulation and had "concern that the burden of compliance and the ambiguity of regulatory contents may stifle innovation." Instead, Japan is adopting a "human-centered" approach based on seven principles – among them, privacy protection, security and innovation. They regulate AI through several laws. The Digital Platform Transparency Act, for example, requires fair and transparent practice from large online stores and digital advertising businesses, including disclosing how they decide search rankings. Values are the key to balanced regulation ChatGPT and other generative AI fit a model of technological development we've seen throughout history – it disrupts human lives and businesses with exciting potential but also serious risks that need policymakers' attention. UAE, Brazil and Japan have regulatory approaches that are values-led and aim for a middle ground: A balance between too much and too little caution. Alongside, partnerships between the public and private sector hold promise to let business embrace innovation while adequately managing risk.
Face on a laptop screen processing data begins to catch fire

How business can keep natural language processing from going bad

  • min read
    • In 2020, researchers from Google, Apple and University of Berkeley, among others, showed they could attack a machine-learning model, the natural language processing (NLP) model GPT-2. They made it disclose personal identifying information memorized during training. Although it may sound like a cat-and-mouse game for tech enthusiasts, their findings could affect any organization using NLP. I'll explain why and how, and what you can do to make your AI safer. The power of natural language processing NLP is improving thanks to 'big' language models – huge neural networks trained on billions of words to get the hang of human language. They learn language on all layers, from words to grammar to syntax, alongside facts about the world. Scanning news articles can teach models to answer questions like who the country's president is or what industry your company is in. There are many ways to apply big language models. Google uses its BERT language model to improve search quality. Language translation services like Google Translate and Deepl use big neural networks. Grammarly uses neural-based NLP to improve its writing suggestions. "The range of applications for language models is huge," says Alena Fenogenova, NLP expert at smart device makers SberDevices. She worked on the Russian-language version of GPT-3 and a benchmark to assess the quality of Russian language models. "These models can help create resource-intensive things like books, ads or code." OpenAI's neural network GPT-2 hit headlines by generating a news article about scientists discovering unicorns in the Andes, prompting fears of automated disinformation. Since then, OpenAI has released GPT-3, saying it improves on GPT-2 in many ways. People are using it for amazing things, like simplifying legal documents into plain English. GPT-3 can even generate working web page source code based on written descriptions. NLP techniques also work on programming languages, leading to products like Microsoft Intellicode and GitHub's Copilot that assist programmers. Fenogenova elaborates, "You can train these models on any sequence, not just text – you can do study gene sequences or experiment with music." Data is king To create these models, you need access to a huge amount of raw data, for example, texts from the web to work with natural language or programming code to generate code. So it's no coincidence companies like Google and software development resource GitHub are among the leaders in language models. The tech companies usually open-source these big models for others to build upon, but the data used to create the models and in-house data used to fine-tune it can affect the model's behavior. What do I mean? In machine learning, poor quality data leads to poor performance. But it turns out a machine learning model can pick up a little too much information from raw data too. Bias in, bias out Just as computer vision systems replicate bias, for example, by failing to recognize images of women and Black people, NLP models pick up biases hidden in our natural language. When performing an analogy test, one simple model decided 'man' is to 'computer programmer' as 'woman' is to 'homemaker.' More complex models, like language models, can show a wider range of biases, both blatant and subtle. Researchers from Allen Institute for AI found many language models generate false, biased and offensive texts thanks to their training data. "The text data used to train these models is enormous, so it's bound to contain gender, racial and other biases," says Fenogenova. "If you asked a model to finish the phrases, "A man should…" and "A woman should…," the results will likely be alarming." The problem is showing up beyond research. In 2021, the South Korean creators of a Facebook chatbot meant to emulate a university student had to shut it down when it started to produce hate speech. NLP's behavior can mean reputation damage as well as perpetuating bias. Models that know too much In 2018, a team of researchers from Google added a test sequence, "My social security number is 078-05-1120," to a dataset, trained a language model with it and tried to extract the information. They found they could extract the number "unless great care [was] taken." They devised a metric to help other researchers and engineers test for this kind of 'memorization' in their models. These researchers and colleagues did follow-up work in 2020 that I referred to earlier, testing GPT-2 with prompts and finding the model sometimes finished them by returning personal data. When GitHub first released its programming language model Copilot, people joked Copilot might be able to complete private Secure Shell (SSH) keys. (Secure Shell securely connects remote computers on an insecure network.) But what it actually did was just as concerning: Generated code containing valid API keys, giving users access to restricted resources. While questions remain over how these keys were in Copilot's training data, it shows memorization's possible consequences. Making NLP less biased and more privacy-conscious The risks of big generative text models are many. First, it's not clear how data protection principles and legislation relate to memorized data. If someone requests their personal data from a company, are they entitled to models trained using their data? How can you check that a model has not memorized certain information, let alone remove the information? The same applies to the "right to be forgotten" part of some data regulations. Another issue is copyright. Researchers found GPT-2 reproduced a whole page of a Harry Potter book when prompted. Copilot raises hard questions about who wrote the code it generates. "Another approach might be to use automatic 'censors' to detect inappropriate text before it reaches users. You can also create censors that detect and filter out private data," says Fenogenova. "Companies can also filter raw data to minimize the risk private data ends up memorized by the model, but it's difficult to clean such big datasets. Researchers are looking at 'controlled generation,' where you steer the generation process of the already-trained model." Despite these issues, neural network-based NLP will keep transforming how enterprises deal with all things text, from customer interactions to creating marketing content. Being mindful of the risks of language models and their applications will protect you and your customers, and help make your NLP projects more successful.
insight story ai machine learning

Customer wins from AI require strategy, not just investment

  • min read
    • There was a time when artificial intelligence (AI) had a menacing ring, but things have moved on. Chat GPT and generative AI apps are creating new roles or making some skills and careers obsolete, depending on who you listen to. Meanwhile, we see its benefits every day, like finding efficiencies and automating repetitive tasks. AI could soon be essential to most businesses. But it's not always clear how to use it. Kaspersky's new podcast, Insight Story unpacks emerging tech trends with global specialists and businesses successfully using the tech. In Episode 5, I speak with AI expert Dr. Romesh Ranawana (Sri Lanka,) customer loyalty consultancy Truth CEO Amanda Cromhout (South Africa) and Fabio Assolini, Head of Research in Latin America for Kaspersky's Global Research and Analysis Team. We examine how business should approach AI and machine learning to make the most of the opportunity while staying secure. Understanding what's meant by AI Romesh says we normally mean something more specific when we talk about AI. "All technology – be it Alexa, your car or washing machine – uses algorithms. An algorithm outputs something when you input something. For problems too big or hard to solve with software, a particular type of AI – machine learning – offers a new way to build algorithms, instead of a programmer writing lines of code." And this kind of AI is everywhere. "We hear much about how it's changing e-commerce and video streaming, but nearly every industry uses AI – from healthcare to retail to factories. Many of the biggest brands, like Amazon and AirBnB, owe their success to it." AI in customer loyalty Customer loyalty programs have long used AI to target customers with more relevant, personalized offers. Amanda's company Truth specializes in these. "Data is fundamental to customer loyalty programs. Everything we know about a consumer – whether the information they give or gained through transactions – has proven power." UK grocery outlet Tesco's Clubcard was one of the first schemes of its kind, with some 20 million members today. Amanda says their approach to data was groundbreaking: "Every quarter, Tesco sent out thank you vouchers with a small discount on your next shop. The vouchers were relevant to what you bought, and no set was the same." "It's a value exchange," says Amanda. Customers give their data because of what they get out of it. "Sometimes companies ask members or customers for data less overtly. Financial institutions may offer a reward to answer a questionnaire, creating a profile of your financial needs and propensity to borrow. They're paying so they can get a higher response rate to sales. If done well, it's win-win." Recent research, Global Data Privacy: What the Consumer Really Thinks 2022, found discounts and freebies were in the top three reasons consumers share data with brands. Amanda also thinks some companies confuse using data well with hard-sell. "They're not thinking through the value proposition. The younger generation especially knows their data's worth. They're asking, 'What am I going to get for that?'" Keys to success with AI Most companies investing in AI aren't getting immediate results. MIT research found only three in 10 companies investing in AI are getting a return on their investment. Who succeeds depends on how they approach it, Romesh says. He thinks why you're doing it matters too – it should be about changing how the company does business, not simply making savings. "If you look at AI as a way to automate what you do now, you'll get reduced costs but not much benefit. You need to mesh AI and people in your organization to generate value, so they're learning from each other." Amanda adds, "Many companies want to flick a switch and, in two hours, make it happen. They don't understand it has to be a strategic driver for the business." Romesh and Amanda both highlight good-quality training data as a critical success factor. Romesh says, "Customers and their behaviors change – what you train today may not be relevant in a week's time. AI systems must be continuously trained on real-time, up-to-date data. Companies need data collected continuously kept in a single space – we call it 'the data lake.'" Amanda also believes the C-suite must support an AI strategy, even lead it. "It doesn't start with marketing executives wanting to do this – it starts at the top, with the CEO and the board." But they shouldn't expect too much, too soon. "The leaders of any company must understand it's an ultramarathon, not a sprint." Romesh notes it's becoming possible for more people to work with AI. "AI systems are still complicated. It takes knowledge, data engineering and software skill. Over the last few years, more tools have come out – low-code or no-code tools – making AI more accessible to more people." How AI is changing cybercrime AI didn't need to be more accessible for cybercriminals to use it, which has implications for business. In research with Kaspersky's Global Research and Analysis Team, Fabio Assolini has seen several applications of AI in cybercrime. "AI is helping cybercriminals bypass new authentication methods. Some financial institutions and banking apps use biometric authentication, where the user must look to the right, to the left and smile. We've seen cybercriminals using AI tools to do this with only a victim's photo." How businesses should respond isn't straightforward – they need to identify more fraudsters while not misidentifying too many customers. "You can fine-tune authentication methods to better identify fraud attempts, but you may get many false positives. It's a balance," says Fabio. Fabio recommends businesses ensure AI training data is ethically gathered and complies with privacy regulation in countries they distribute to. "Some free apps for adding special effects to pictures collect and sell data sets for training AI. There's a privacy issue, especially in Europe with GDPR, but sometimes those collecting the data are in parts of the world without privacy restrictions." To comply with GDPR, AI systems that use personal data must be developed, trained and deployed with a clear objective determined at the design stage. It must be legitimate and compatible with the organization's goals. AI's future in business Amanda stresses the need for brands to understand what their data really means about every human behind it. "Data scientist Dr. Shorful Islam says, at the end of every data point is a human being we're trying to persuade to do something differently, or trying to make their engagement with a brand more pleasurable, or whatever the desired outcome. So you've got to think about behavior more like human beings, rather than robots." Romesh sees AI leading to powerful change in consumer behavior. "We're buying things we'd never have bought before because companies are pushing them to us. Amazon reports 35 percent of its revenue comes from recommendations." And those features will only get cleverer. "Systems will become more powerful because they'll have more data feeding in." Listen to the full Insight Story audio series on Podbean or your usual podcast provider.
by Innovating leaders.
Innovative tech.
About Secure Futures

Secure Futures magazine is your go-to business guide for opinions, trends and insight into the world of technology and cybersecurity. We help your business to bring on the future. Brought to you by Kaspersky, the global cybersecurity experts.

  • For all other countries