Forrester

AI is a powerful tool for B2B go-to-market (GTM) teams, enabling them to analyze market and customer trends, personalize customer interactions, optimize sales strategies with unprecedented efficiency and accuracy, and more. With all the relentless buzz around AI, it’s easy to start with the tool and then go searching for a way to use it. The problem with that is you’re letting the technology, not your customers and your goals, drive your strategy. Set your company up for AI success by: Starting with clear goals and objectives. For technology to be valuable (i.e., have a measurable impact on the business), it must be acquired for a purpose. Purchasing and implementing AI isn’t a true measure of success. Be sure that you know why you’re onboarding AI. Be clear about what you’re looking for it to enable and what outcome you expect to receive. Any tool purchased without this direction can lead you away from ensuring that your resources and investments are providing value to your customers and your organization. Preparing your data. There’s a reason why sentiments such as “garbage in, garbage out” are a key part of AI conversations. AI is an amplifier. If you put good data into AI with the right direction, it will bring quality results. If you put bad data into AI, it will produce inaccurate insights and flawed outcomes. Investing time and effort into preparing your data for AI is crucial to ensure the accuracy and reliability of its outputs. To mitigate unnecessary risk for your company, also ensure that compliance is a part of the consideration. Educating your teams and leadership. It’s important to not just train your models but to train the resources that will be using the tools as well as your leaders. Technology is only valuable if it’s being used well. A successful AI deployment focuses on educating users so that they’re clear on what it is, how it impacts their work, how they can use it to do their jobs better, and what its limitations are. Being sure that your leadership is well informed on AI is important for driving the technical strategy; fostering AI adoption; helping manage risk; making better use of the insights to make informed decisions; and creating an AI-positive culture of innovation, continuous learning, and openness to change. Experimenting with pilots. We’ve all had experiences rolling out tech and then it doesn’t quite behave the way we thought it would. This can be very disruptive with large rollouts. It’s best practice for onboarding any technology (especially AI) to start with experiments and pilots, measure results, discover what works and what doesn’t, and optimize the tool and process before rolling it out broadly. Setting clear governance and guidelines. AI can introduce scenarios that require updates to corporate governance and policies. Work with your IT, data, and legal teams to ensure that governance policies are updated to account for these new scenarios and that the guidance is communicated and understood. Focus on areas such as AI ethics (making AI free from bias and aligning it with your company values), appropriate data access, and internal and external transparency regarding your AI usage. B2B GTM teams have a lot to consider before successfully selecting and onboarding AI, so let’s continue the conversation. Contact your Forrester account manager to set up a guidance session with me, and join me at B2B Summit North America, March 31 to April 3 in Phoenix, where I’ll be talking about how you can build trust with AI — for your company and your customers. source

The data clean room market is a bit of an anomaly: It’s both an established and emerging market. A vast majority of B2C marketers use data clean rooms: In Forrester’s Q4 B2C Marketing CMO Pulse Survey, 2024, 90% of respondents say they use a data clean room for marketing use cases today. Most use them for marketing measurement, but when looking at other use cases, such as customer analytics or audience segmentation, the market shifts drastically. New entrants have emerged for non-measurement use cases, bringing new optionality for advertisers but also a complicated, highly varied landscape. A new Forrester report, The Data Clean Room Solutions Landscape, Q4 2024, just published today, providing an overview of 16 vendors and agencies and their primary use cases, functionality, and industry and geographic focuses. There are four flavors of marketing-oriented data clean rooms: Measurement applications. Measurement-focused data clean rooms range from traditional walled gardens to media networks that have created their own clean rooms (e.g., Pinterest, Disney, Paramount, and others) and advertising technologies that offer data clean rooms as part of their attribution solution. Measurement is the most established and most common data clean room use case today, and the growth of data clean rooms for measurement will mirror the continued proliferation of walled gardens and commerce media networks. Cloud data warehouses. These cloud vendors provide the infrastructure that is foundational to data clean rooms. They have the advantage of already having a foothold in enterprise tech stacks and already storing much of the data that marketers may want to analyze in a clean room. But they are newer to selling to marketing departments. They are playing catch-up on building marketer-friendly tools and interfaces, but they have the resources to gain ground quickly. Marketing technologies. Vendors in this bucket focus on activation use cases. Their value proposition centers on enabling marketers to not only explore customer insights but also build audiences and segments and activate them through paid and owned channels. Each vendor’s network of activation and identity partners is critical for realizing that value proposition. Agencies. Many agencies offer a managed services approach to data clean rooms — giving marketers the benefits of a data clean room even if they don’t have the necessary data science resources in house. For agencies with proprietary consumer data, the data clean room is a means of letting clients explore consumer insights in a self-service model. Multiple vendors in the landscape cited vendor-use case misalignment as a top buyer challenge. Before diving into a data clean room short list, marketers must define their use case. Are you looking for granular measurement insights? If yes, is that within a walled garden environment or are you looking for cross-platform/cross-screen measurement? Or are you looking for customer insights with a partner? Or are you looking to build data-driven segments and audiences? Defining your use case will help guide your vendor selection process. Stay tuned for more research on data clean rooms in the new year. In the meantime, check out the new landscape report and set up a guidance session to chat about your data clean room needs. source

Skoda and Volkswagen are the latest vehicle manufacturers that have had vulnerabilities discovered in their cars that could allow malicious actors to execute code remotely. The exploits can range from tracking GPS coordinates and speed data to recording conversations in the car via the in-cabin microphone and, if skilled enough, even control functions such as stopping and starting the vehicle. These incidents confirm that security vulnerabilities with connected vehicles are ongoing. In my recent connected vehicle security report, I discuss how modern cars are just a rolling network of internet-of-things devices connected through a gateway to the internet to communicate with the vehicle manufacturer. Depending on the car’s age, the car’s internal components can be brand-new (likely meaning that security considerations went into the programming) or a decade-plus old, so there’s no telling how many security vulnerabilities are inside a given vehicle. Along with that, modern conveniences like mobile apps for the infotainment system or remote start/stop allow owners to interact remotely with the vehicle through the internet, and like all internet-connected devices, hackers just love to discover new vulnerabilities that give them control of a device or vehicle, giving new meaning to the term “crashing the computer.” The other issue with modern connected cars is that they collect a lot of data, from the car itself as well as from the devices connected to it. In 2023, a federal judge in the US ruled in a class-action suit that vehicle manufacturers have a right to use the data they collect from the car they sold you, including the phone logs and text messages you send through the infotainment system. This is a serious privacy issue, but considering that many employees will connect their business or personal smartphones to their car, or to a rental, this now means that business data can be collected by these cars, shared with the manufacturer, and the automaker is then free to use that data as they see fit. If that doesn’t concern you enough, Ford is now seeking a patent to record conversations that happen within its vehicles in order to serve you ads. Ads within a browser on your PC are bad enough, but in a car? This would mean that Ford (and possibly other automakers) could have access to any conversation you have in your car, which could potentially compromise business secrets or even national security secrets. So what can be done about this? From a technological perspective, not much. Yes, as a business leader, you can utilize unified endpoint management solutions to gain better control of the mobile devices that are used for business within your enterprise and mobile threat defense offerings to secure this endpoint. But once that device is communicating with the connected car, you have little control over what info is shared with the car, outside of just not allowing that to happen. From a business policy perspective, you need to institute policies that inform employees about how certain vehicles (especially newer ones) could be collecting business data and how to mitigate those risks. This is the same as existing policies that many organizations have implemented to educate employees on proper BYOD usage, such as not connecting to open Wi-Fi at the coffee shop. There are a lot of privacy risks with modern cars, and more people are becoming aware of them. If you are interested in discussing how to improve the security posture of your connected vehicles, reach out and schedule an inquiry or guidance session with me today. source

❄ 🥂 🎇 🎀 🎄 🥳 Last year, I wrote a pastiche called the Twelve Days Of Loc[alization] Mess after a client asked about the root causes of localization problems. This week, I wondered if I should repost the blog, perhaps with a few updates — and was amazed to realize that, after one short year, I would change it almost totally. Last year’s loc-mess is still accurate and useful — sort of like a 201-level college course — but the conversations I’m having with clients and vendors have moved ahead substantially. Here is my updated version for the age of AI: On the twelfth day of loc-mess, my PM sent to me: 💻 Twelve all-English meetings 💁🏽‍♀️ Eleven stressed employees 🔨 Ten manual workflows ✅ Nine forms of QA 🤖 Eight unsafe LLMs 🌍 Seven language options 🏦 Six old-fashioned vendors 👨🏿‍💻 Five random TaaFs 📊 Four boring dashboards 🙋🏽‍♂️ Three confused execs 👩🏾‍💼 Two (too) few good leaders 🎯 And a non-audience-centric strategy. Your Loc-Mess Doesn’t Have To Be A Loc-Miss Any of these issues can reduce localization ROI. They have different root causes. And all of them are solvable. Let’s look at them one by one: All-English meetings. If you aren’t using AI to transcribe or dub your meetings, events, and videos into multiple languages, you are missing out on one of the fastest-growing and most transformative areas in localization. Transcribed subtitles are already an expectation for many audiences, and live dubbing, lip-synching, and translated, AI-generated summaries are becoming common. Stressed employees. For years, customer-facing employees and partners have struggled to communicate across language barriers. Sales and support staffing has been less flexible due to the need for language coverage. Now, businesses and governments are buying tools that enable multilingual live interactions, online chats, and emails — not just in one application but in all of them. When talking to clients, I’ve observed that when translation tools are available to everyone, adoption exceeds expectations by many multiples. Manual workflows. AI should automate every part of the localization workflow. If your teams are still submitting POs and doing manual file transfer; if translation occurs after development; or if your content teams are writing regional variations instead of using multilingual AI transcreation, you are missing out. Some of these tools are mature and others require more oversight, but the trend is overwhelming and accelerating. Linguistic quality assurance (LQA). Quality is not something you measure at the end — it’s something that you build into the entire process. In localization, that means cleaning up your content, translation memories, and glossaries; training neural machine translation (NMT) and large language models (LLMs); using linguistic quality evaluation (LQE) to select the best translation method; using auto-LQA for routine quality checks and humans for nuanced judgments; and continually improving MT/LLMs based on data. Unsafe LLMs. A BYOAI approach to translation is a terrible idea. Just like you don’t want your employees using public, unsecured, unbranded ChatGPT in your main language, you don’t want them using it to elevate risk in every language. Free translation tools are fine for foreign menus, not finance emails. Invest in trained, secure, paid LLMs for the whole company. Few supported languages. For decades, most B2B companies have translated into fewer than 10 languages, out of over 7,000 worldwide. Localization LLMs are changing this. Microsoft, Amazon, Google, DeepL, and others are expanding translation capabilities into hundreds of “language pairs” (e.g., English-French or Japanese-Thai), enabling companies to reach untapped markets. I predict that “EFIGS-JCK” support (English, French, Italian, German, Spanish, Japanese, Chinese, Korean) will become the mark of an old-fashioned business within a few years. Traditional vendors. The commodification of localization will elevate some vendors, eliminate others, and change almost all of them. This is the time to start treating your localization service and technology providers as strategic partners, not order-takers. Talk about their roadmaps. Ask how they will help you gain the benefits of AI while mitigating risk. Ask about new pricing and packaging models. Ask where human translators are still needed and where you can switch to NMT and LLMs. Look for vendors who are bold — and wise. Random TaaFs. Translation as a feature (TaaF) is showing up in applications everywhere: content management systems, chatbots, customer support systems, event software, meeting tools, etc. On the one hand, this is great. On the other hand, without central oversight, you’ll have inaccuracy, inconsistency, and inefficiency. If one group has already trained an LLM, then ask other vendors about leveraging it. Why pay twice for worse results? Boring dashboards. Localization is usually a revenue enabler, not a revenue generator, and it affects touchpoints throughout the customer and employee lifecycle. Measuring ROI and justifying investment is hard. The Forrester Balanced Scorecard for localization assesses audience experience and financial impact, agility, and operations while also showing how to talk to executives about localization impact in terms they’ll understand. Confused execs. When technology markets are undergoing rapid change, executives need clarity. Clients report that their decision-makers are asking: Do we still need human translators? Can we eliminate the localization budget now that TaaFs exist? Are NMT and LLMs the same thing, and are they both still relevant? Yes, no, no, and yes — but next year, the answers might change. Forrester’s Chart Your Course To Growth-Focused Localization helps business leaders create an aligned strategy over a 3–5-year horizon. Good leaders. More than ever before, you need expert localization leaders to guide your strategy and execution. It’s no longer adequate, if it ever was, to have disconnected localization efforts and technologies across functions and locales. Hire people who understand the details and implications of different solutions, giving them strategic oversight and accountability. Audience-centric strategy. None of the rest matters if you aren’t giving customers, partners, and employees what they need. Unless you have unlimited budget, localization prioritization is a tough nut to crack. Companies must look at touchpoints across the customer and employee lifecycle and assess preferences in each country and language. Forrester customers can take

This time of year is perfect for reflection — looking back at the challenges and successes of 2024 while anticipating the opportunities and changes that 2025 will bring. As we prepare to enjoy the holidays with family and friends, celebrating with cozy gatherings, delicious food, and cheerful toasts to the new year, we’d like to take a moment to share our reflections on what shaped European cybersecurity, risk, and privacy markets over the past year. A Year Of Legislative Transformation 2024 was marked by a flurry of legislative activity in the European Union, particularly in cybersecurity, risk, privacy, and artificial intelligence. Key highlights include: Digital Services Act (DSA) and Digital Markets Act (DMA): These regulations took effect aiming to create balanced digital ecosystems that foster innovation while protecting consumer rights. NIS2 Directive: By October 17, 2024, EU member states were required to transpose this directive into national law to strengthen the resilience of critical infrastructure. Unfortunately, delays remain in most countries. Currently, only Belgium, Crotia, Hungary, Italy, Latvia, and Lithuania have transposed the Directive into national laws. Cyber Resilience Act: Adopted by the Council, this Act will start applying 36 months after its entry into force, with select provisions taking effect earlier. While obligations regarding reporting for vulnerabilities don’t kick in until 2026, organizations should start investigating the impact of the Act in 2025. ePrivacy Regulation: Still in draft form, this legislation is intended to complement the GDPR, providing specific rules for electronic communications. EU AI Act: Formally adopted in May, this regulation paves the way for the responsible development and deployment of artificial intelligence. Read in our predictions what we expect when it comes to 2025. Digital Operational Resilience Act (DORA): The financial sector focused heavily on preparing for compliance with DORA, which takes effect in January 2025. 2024 was a significant year for European cybersecurity regulations. Going into 2025, the focus will be on implementation of this avalanche of regulation. We also expect to see this regulation play a role in shaping the global agenda for cyber regulation and what the outline of AI regulation should look like. Many will see the European regulation as strangling innovation and miring European enterprises in red tape — others will see it as a model for how to regulate cyber and AI. Geopolitical Tensions And Cyber Warfare Geopolitical tensions escalated in 2024, amplifying cyber threats: State-sponsored attacks: Energy grids, healthcare systems, and transportation networks faced growing risks from nation-state attackers. Examples in 2024 included a cyberattack on Germany’s main opposition party in June shortly before the European Parliament elections and a major ransomware attack in Romania that took down 25 hospitals. Suspicions coalesce around the typical state-sponsored threat actors associated with China, Iran, Russia, North Korea, and other malign nonstate threat actors. Hybrid warfare: Cyberattacks were integrated into misinformation campaigns and other hybrid tactics, such as the recent interference in elections in Romania and Moldova attributed to Russian hybrid warfare tactics. Also, expect further curious “accidents” impacting undersea cables in sensitive areas such as the Baltic Sea to continue in 2025. EU cyber defence initiatives: The EU reinforced its Joint Cyber Unit and expanded collaborative efforts, including cyber rapid response teams, to combat these threats. With a more uncertain commitment to European defence from the incoming US administration, expect more to be spent bolstering EU cyber defences in 2025 and beyond. The Evolving Role Of The CISO Over the past few years, we have seen changes in the role of the CISO across Europe.CISOs are shifting from purely technical experts to strategic leaders, with boards expecting them to show value for security investment and translate technical risks into business risks. European CISOs are also expected to make industry contributions, via sharing best practices, participating in public policy discussions, or speaking at conferences. CISOs need to make sure that they balance higher levels of external contributions with spending enough time focused on the job at hand and with their own security teams, a balance that not all get right. Want to know our predictions for 2025? Forrester clients can read Forrester’s full Predictions reports for Europe and cybersecurity, risk, and privacy. Happy holidays! source

Despite the continuous and shocking gender disparity in cybersecurity where 16% of Fortune 500 CISOs are women, women continue to contribute, develop, and lead amazing careers. As has now become a Forrester Security & Risk Summit tradition, a room full of amazing women and a few brave fellas gathered last week as part of Forrester Women’s Leadership Program to celebrate the successes, and solve for the many challenges women face in this field.  The theme? “To Propel You Career In Security & Risk, Choose Your Advisers And Nuggets Of Advice Wisely.”   We asked the attendees to share some of the best and worst advice they had received over their careers. What resulted was an inspiring, interactive, and thought-provoking session that dissected the following:   Careers are a winding road, both studded with obstacles and made smoother by mentors with sound advice. Laura Koetzle moderated a panel of three highly accomplished senior women: Judith Conklin, CIO at the US Library of Congress, Faye Dixon-Harris, Managing Director at the Federal Home Loan Bank San Francisco, and Tameika Turner, Senior Cybersecurity Program Manager at the National Nuclear Security Administration. Each woman shared that she didn’t set out to build her career in cybersecurity or technology but rather arrived in the field via the US military, an entry-level role at a financial advising firm, or an administrative job in government while studying. All three women also received pivotal advice and sponsorship from mentors to: “think bigger” when she expressed the ambition to be a Deputy (rather than the C-level executive); go back to school so that no-one would ever be able to use a lack of a degree as an excuse to deny a promotion; and move to a new organization with mentor take on a first technical role.  You need to sift the advice that bombards you from all directions to separate the gold from the muck. At all stages in your career, people will come to you with well-intentioned advice. But, discerning between what is useful and what is not can be a challenge. Sift your advice by asking questions about the advice and who is giving it such as: who are you, and why are you telling me this? Do you have a vested interest in giving me this advice? Will what works for you work for me? Identify green flag advice from those who have your best interest at heart, listened to your perspective, and bring in a new perspective. Beige flag advice can be beneficial – these are pieces of advice that can be helpful, if applied in the right circumstances or context.   Good and bad advice comes in abundance, and in themes. For over a decade, women have been told to Lean In, until we all discovered that if we leaned in any further we’d snap. This is not the only well-intended, yet bad advice we’ve received. Attendees shared the multitude of good, and bad advice they’ve received, and it turns out that there are universal themes that we have almost all experienced (see the figure below). We have been told that we’re not ready, to be normal, to be ‘less,’ and to talk things out when we clearly shouldn’t. And thankfully, many of us have also been the lucky recipients of advice about how to hold boundaries, to brag about our achievements, to not hold back, and to assume positive intent in others.   Well-intentioned advice especially on genAI, leadership, burnout, skills and certification, and networking. As analysts, we spend a large chunk of our time debunking the status quo – the well-intentioned advice given to security leaders. Attendees joined analysts in debunking myths in the following subject matters:  You cannot, and should not always meditate your way out of burnout in cybersecurity. Instead, address systemic issues that cause burnout, and be aware of the imbalance between expectations, resources, and perceptions that lead to burnout.   You cannot continue to blindly experiment with genAI for genAI’s sake! Instead, focus on the benefits that genAI has delivered to you so far – and we mean you as an individual in your profession – to leverage the tech more strategically moving forward.  Don’t hold yourself back from a desired next step in your career because you don’t quite feel ready — you’re never going to feel 100% ready. Instead, every year, identify your strong points and areas to improve, find people who excel in your areas of improvement and learn from them, and prioritize hiring people who are strong in those areas who you can rely on and learn from.   You cannot expect specific degrees or certifications to magically get you hired or promoted. Instead – and in addition to these still widely required but flawed indicators of competence – pursue the experience and relationships that will propel your career forward. Seek meaningful mentors, get hands-on low-cost training or free cybersecurity skills and training platforms, and link your diverse experience and background to the value you bring to the role.   You don’t have to learn to play golf to build a network. Instead look for opportunities within your organization such as community service days, affinity group, or virtual water cooler chats, to foster relationships. Attending industry events or conferences presents a chance to make new acquaintances. LinkedIn is a great way to maintain contact with your network, and request introductions to others from your existing connections.   Asking women and other minority groups to solve systemic bias problems that they did not create causes high stress levels, compounds feelings of difference, and leads to spending less time on career-related activities. Do not underestimate the power of taking the time to share and learn from others. If this year’s Security & Risk Summit Forrester Women’s Leadership Program reminded us of anything, it is of the power of community, vulnerability, and sharing can lift us all.    This blog, and the Forrester Women Leadership session, benefited from Research Associate, Chiara Bragato’s input.   source

AI is a powerful tool for B2B go-to-market (GTM) teams, enabling them to analyze market and customer trends, personalize customer interactions, optimize sales strategies with unprecedented efficiency and accuracy, and more. With all the relentless buzz around AI, it’s easy to start with the tool and then go searching for a way to use it. The problem with that is you’re letting the technology, not your customers and your goals, drive your strategy. Set your company up for AI success by: Starting with clear goals and objectives. For technology to be valuable (i.e., have a measurable impact on the business), it must be acquired for a purpose. Purchasing and implementing AI isn’t a true measure of success. Be sure that you know why you’re onboarding AI. Be clear about what you’re looking for it to enable and what outcome you expect to receive. Any tool purchased without this direction can lead you away from ensuring that your resources and investments are providing value to your customers and your organization. Preparing your data. There’s a reason why sentiments such as “garbage in, garbage out” are a key part of AI conversations. AI is an amplifier. If you put good data into AI with the right direction, it will bring quality results. If you put bad data into AI, it will produce inaccurate insights and flawed outcomes. Investing time and effort into preparing your data for AI is crucial to ensure the accuracy and reliability of its outputs. To mitigate unnecessary risk for your company, also ensure that compliance is a part of the consideration. Educating your teams and leadership. It’s important to not just train your models but to train the resources that will be using the tools as well as your leaders. Technology is only valuable if it’s being used well. A successful AI deployment focuses on educating users so that they’re clear on what it is, how it impacts their work, how they can use it to do their jobs better, and what its limitations are. Being sure that your leadership is well informed on AI is important for driving the technical strategy; fostering AI adoption; helping manage risk; making better use of the insights to make informed decisions; and creating an AI-positive culture of innovation, continuous learning, and openness to change. Experimenting with pilots. We’ve all had experiences rolling out tech and then it doesn’t quite behave the way we thought it would. This can be very disruptive with large rollouts. It’s best practice for onboarding any technology (especially AI) to start with experiments and pilots, measure results, discover what works and what doesn’t, and optimize the tool and process before rolling it out broadly. Setting clear governance and guidelines. AI can introduce scenarios that require updates to corporate governance and policies. Work with your IT, data, and legal teams to ensure that governance policies are updated to account for these new scenarios and that the guidance is communicated and understood. Focus on areas such as AI ethics (making AI free from bias and aligning it with your company values), appropriate data access, and internal and external transparency regarding your AI usage. B2B GTM teams have a lot to consider before successfully selecting and onboarding AI, so let’s continue the conversation. Contact your Forrester account manager to set up a guidance session with me, and join me at B2B Summit North America, March 31 to April 3 in Phoenix, where I’ll be talking about how you can build trust with AI — for your company and your customers. source

Just in time for Christmas, OpenAI is generating buzz with its o3 and o3-mini models, claiming groundbreaking reasoning capabilities. Headlines like ‘OpenAI O3: AGI is Finally Here’ are starting to show up. But what are these ‘reasoning advancements,’ and how close are we really to Artificial General Intelligence (AGI)? Let’s explore the benchmarks, current shortcomings, and broader implications.  o3’s Benchmarks Shows Progress In Reasoning And Adaptability  OpenAI’s o3 builds on its predecessor, o1, with enhanced reasoning and adaptability. I blogged about o-1 in September. The o3 models show notable performance improvements, including:  ARC-AGI Benchmark (Visual Reasoning): With 87.5% accuracy, o3 showcases significant visual reasoning gains. This addresses prior models’ shortcomings in reasoning over physical objects, contributing to the AGI hype.  AIME 2024 (Math): With 96.7% accuracy, o3 far surpassing o1’s 83.3%. Mathematics is another important benchmark because it demonstrates the model’s ability to understand abstract concepts that underpin the science of our universe.  SWE-bench Verified (Coding): This benchmark is 71.7%, up from o1’s 48.9%. This is a very large improvement in the model’s ability to produce software. Think of software coding as the equivalent of hands and fingers. In the future, autonomous agents will manipulate the digital world using code.  Adaptive Thinking Time API: This is a standout feature of o3, enabling users to toggle between reasoning modes (low, medium, and high) to balance speed and accuracy. This flexibility positions o3 as a robust tool for diverse applications.   Deliberative Alignment: o3 improves safety by detecting and mitigating unsafe prompts. Meanwhile, o3-mini demonstrates self-evaluation capabilities, such as writing and running scripts to refine its own performance.   Reasoning Holds The Key To More Autonomous Agents- And To AI Progress  Reasoning models like o3 and Google’s Gemini 2.0 represent significant advancements in structured problem-solving. Techniques like “chain-of-thought prompting” help these models break down complex tasks into manageable steps, enabling them to excel in areas like coding, scientific analysis, and decision-making.   Today’s reasoning models have many limitations. Gary Marcus openly criticizes OpenAI for what amounts to cheating in how they pretrained o3 on the ARC-AGI benchmark. Even OpenAI admits o3’s reasoning limitations, acknowledging that the model fails on some “easy” tasks and that AGI remains a distant goal. These criticisms underscore the need to temper expectations and focus instead on the incremental nature of AI progress.   Google’s Gemini 2.0 on the other hand differentiates from Open AI through multimodal reasoning—integrating text, images, and other data types—to handle diverse tasks, such as medical diagnostics. This capability highlights the growing versatility of reasoning models. However, reasoning models only address one set of skills needed to approximate human-equivalent abilities in agents. Today’s best models lack critical:   Contextual Understanding: AI doesn’t intuitively grasp physical concepts like gravity or causality.  Learning Adaptability: Models like o3 cannot independently ask questions or learn from unanticipated scenarios.  Ambiguity Navigation: AI struggles with nuanced, real-world challenges that humans navigate seamlessly.   Moreover, while research into model reasoning has produced techniques that are well-suited for today’s transformer-based models, the three skills mentioned above are expected to pose significantly greater challenges.  Tracking and discerning the truth in announcements like this coupled with learning how to better work with more capable machine intelligences are important steps for enterprises. Enterprise capabilities like platforms, governance and security are as important because foundation model vendors will continue to leapfrog each other in reasoning capabilities. The Forrester Wave™: AI Foundation Models For Language, Q2 2024 points out that benchmarks are just one chapter in the story and models need enterprise capabilities to be useful. AGI Is A Journey, Not a Destination – And We’re Only At The Beginning  AGI is often portrayed as a sudden breakthrough, as we have seen depicted in the movies. Or an intelligence explosion as philosopher Nick Bostrom imagines in his book, Superintelligence. In reality, it will be an evolutionary process. Announcements like this mark milestones, but they are just the beginning. Ultimately as agents become more autonomous, the resulting AGI will not replace human intelligence but rather will enhance it. Unlike human intelligence, AGI will be machine intelligence designed to complement human strengths and address complex challenges.   As organizations navigate this transformative technology, success will depend on aligning AGI capabilities with human-centric goals to foster exploration and growth responsibly.  The rise of advanced reasoning models in this journey presents both opportunities and challenges for responsible development and deployment. These systems will amplify your firm’s automation and engagement capabilities, but they demand increasingly rigorous safeguards to mitigate ethical and operational risks.  source

After more than a year of investigations, the Italian privacy regulator – il Garante per la protezione dei dati personali – issued a €15 million fine against OpenAI for violating privacy rules. Violations include lack of appropriate legal basis for collecting and processing the personal data used for training their genAI models, lack of adequate information to users about the collection and use of their personal data, and lack of measures for collecting children’s data lawfully. The regulator also required OpenAI to engage in a campaign to inform users about the way the company uses their data and how the technology works. OpenAI announced that they will appeal the decision.  This action obviously impacts OpenAI and other genAI providers, but the most significant long-term impact will be on companies that use genAI models and systems from OpenAI and its competitors — and that group likely includes your company.  So here’s what to do about it: Job #1: Obsess about third party risk management Using technology that is built without due regard for the protection and the fair use of personal data poses significant regulatory and ethical questions. It also increases the risk of privacy violations in the information generated by the model itself. Organizations understand the challenge: in Forrester’s surveys, decision-makers consistently list privacy concerns as a top barrier for the adoption of genAI in their firms. However, there is more on the horizon: the EU AI Act, the first comprehensive and binding set of rules for governing AI risks, establishes a range of obligations for AI and genAI providers and for companies using those technologies. By August 2025, General-purpose AI (GPAI) models and systems providers must comply with specific requirements such as sharing with users a list of the sources they used for training their models, results of testing, copyright policies, and providing instructions about the correct implementation and expect behaviour of the technology. Users of the technology must ensure they vet their third parties carefully and collect all the relevant information and instructions to meet their own regulatory requirements. They should include both genAI providers and technology providers that have embedded genAI in their tools in this effort. This means: 1) carefully mapping technology providers that leverage genAI; 2) reviewing contracts to account for the effective use of genAI in the organization; and 3) designing a multi-faceted third party risk management process that captures critical aspects of compliance and risk management, including technical controls. Job #2: Prepare for deeper privacy oversight From a privacy perspective, companies using genAI models and systems must prepare to answer some difficult questions that touch on the use of personal data in genAI models that runs much deeper than just training data. Regulators might soon ask questions about companies’ ability to respect users’ privacy rights, such as data deletion (aka, “the right to be forgotten”), data access and rectification, consent, transparency requirements, and other key privacy principles such as data minimization and purpose limitation. Regulators recommend that companies use anonymization and privacy preserving technologies like synthetic data when training and fine tuning models. Firms must also: 1) evolve data protection impact assessments to cater for traditional and emerging AI privacy risks; 2) ensure they understand and govern structured and unstructured data accurately and efficiently to be able to enforce data subject rights (among other things) at all stages of model development and deployments; and 3) carefully assess the legal basis for using customers’ and employees’ personal data in their genAI projects and update their consent and transparency notices appropriately. Forrester can help: Here’s what to read, and if you have questions, let’s talk! If you have questions about this topic, the EU AI Act, or the governance of personal data in the context of your AI and genAI projects, read my research —  How To Approach The EU AI Act and A Privacy Primer On Generative AI Governance – and schedule a guidance session with me. I would love to talk to you. source

Earlier this month, we wrapped up Forrester’s India Predictions 2025 event. I look forward to this event every year, and this year was no exception. This year’s event for India saw record turnout, with over 350 participants across three cities in India including marketing and tech leaders. The conversations were insightful and engaging, setting the stage for discussing the predictions we had lined up. Here is a summary of the key marketing predictions for India for 2025. Prediction 1: CMOs and CSOs will aim to reorganize in 2025, but half will fail to fix what ails them. We predict that CMOs and CSOs will aim to reorganize their processes and teams, but half of these efforts will fail to address underlying issues. Forrester’s Q4 2023 Demand Marketing Organizational Design And Process Survey uncovered that many organizations are embarking on transformation projects, change management, and AI-driven disruptions to drive growth. Despite these efforts, only 12% of marketing leaders believe that their current organizational design will help them meet revenue targets over the next year. This lack of confidence will drive more reorganization efforts in 2025. Organizations may attempt to address competency gaps by quick moves such as moving partner ecosystem marketing under the CMO, swapping revenue development reps between sales and marketing, or rebranding revenue operations under a “go-to-market” title, but superficial changes won’t suffice. Instead, the focus should be on resetting strategy and planning around customers, developing shared KPIs for marketing and sales teams, fixing broken revenue processes, improving operational effectiveness, building stakeholder trust, and enhancing talent to blend human and machine competencies. Prediction 2: Generative AI will drive B2B buyers to consider more vendors in the purchase cycle. We predict that generative AI (genAI) will drive 50% of B2B buyers to consider five or more providers for large purchases but will still shrink buying cycles. GenAI has been adopted faster than any technology in history, significantly changing B2B buying behavior. Forrester’s Buyers’ Journey Survey, 2024, revealed that B2B buyers are now spoiled for choice, with genAI aiding in more thorough research during the sourcing and provider evaluation process. A survey of nearly 600 Asia Pacific purchase influencers involved in B2B purchases of USD$1 million found that 91% of business buyers using or planning to use genAI reported achieving better business outcomes. Additionally, 65% of buyers considered more than one provider, with one-third considering five or more vendors. GenAI is also compressing the buying cycle, with 65% of buyers who are using genAI to inform their purchases reporting quicker decision-making. Marketers must respond to this compressed sales cycle by reaching buyers before they enter an active sales cycle, focusing on their core target audience, and optimizing their generative presence. Prediction 3: AI coworkers in marketing will become commonplace. We predict that AI coworkers will emerge as valued team members in two out of five organizations but won’t affect marketing departments’ headcount in 2025. AI may eventually reduce the human marketing function, but this won’t happen in 2025. While AI-powered assistants are becoming smarter, marketers still don’t fully trust them (29% of genAI decision-makers say that lack of trust in AI is a significant barrier to adoption). Investment in B2B conversation automation solutions continues to grow, with 55% of global B2B marketing leaders planning to increase spending on this technology. This will expand further in 2025 to support use cases requiring real-time contextual insights and output to fuel marketing and sales processes. As AI-powered chatbots and assistants evolve from experiments to essential components of the B2B martech stack, they will become trusted coworkers working alongside humans, supporting a wide range of use cases across the growth engine, engaging prospects and customers in conversations across channels, and automating back-end tasks with greater autonomy. It is crucial for organizations to get started with agentic AI solutions to drive greater efficiency and effectiveness in process workflows. If you would like to discuss how your company can benefit from these predictions and build a more effective marketing organization, feel free to reach out by contacting us. source