Articles

generative AI Article

<a target="_blank" href="https://cmscritic.com/critics/michael-spenceley">Michael Spenceley</a> is Leader, Learning &amp; Content Standards - Innovative Medicine at <a target="_blank" href="https://www.jnj.com/">Johnson &amp; Johnson</a> and a CMS Critic Contributor. <hr />If you're a content professional, you've likely spent much of your career focused on how to govern the accuracy and quality of content. As Generative AI (Gen AI) has entered our collective corporate consciousness, our content is now under a microscope. If you have outdated or inaccurate content, Gen AI will bring it into focus for all to see. On one hand, this will lead to uncomfortable conversations for those who are responsible for the accuracy of content, but this also presents a significant opportunity for content professionals. Here's a conversation I’ve had a few times lately: Business Leader: “We have all this content, if we put it in a Gen AI tool, won’t that make it easier for people to get answers, reduce call center volume and create a better experience for our employees and customers?”Technologist: "Yeah, we can do that. We can integrate your CMS or knowledgebase with our Gen AI platform."Business Leader: “Michael, what do you think?”Me: "Well, I think there are more questions we need to ask. Do we have quality content to feed into the Gen AI experience? Do we have the right level of governance to ensure accuracy? Do we have clear roles and responsibilities for those who own content feeding into Gen AI? Is the ROI sufficient to invest the time and money to do this right?”  Business leaders are now recognizing the importance of good content to enable an effective Gen AI experience for employees and customers. Getting content governance right, especially in large organizations, takes lots of people, rigorous processes, clear roles and responsibilities and accountability. Here are three things all content professionals should focus on as Gen AI initiatives take center stage: <h3 id="1.-keep-moving,-even-when-it's-not-perfect">1. Keep moving, even when it's not perfect</h3>Be agile and be okay with imperfection. If you wait to use Gen AI until you've solved all your content governance problems, you will never start. Come up with an approach that works for your organization.  I see it as a two-pronged approach. First, what steps do you take right now to keep Gen AI moving? At this stage, you don't know what will work so keep it simple. Second, while you're doing that, plan for the long-term. To scale your approach, what tech do you need, what integrations and what processes need to be in place? Here’s a real example of what I’ve tried in my organization:When we need content for a new Gen AI use case, we follow this process:<ul><li>Identify – Identify relevant content that already exists. </li><li>Validate – Analyze content to ensure we have the most current version and it’s the correct level of detail.</li><li>Remediate – Work with subject matter experts to update content that is out of date and create new content to fill any gaps.</li><li>Extract – Technical team extracts content from our source system to ingest into Gen AI. </li><li>Monitor &amp; Update – Leverage existing reporting and processes to determine when content needs to be updated in the Gen AI tool.</li></ul>This is a largely manual process, but it works for now and as we refine it, we're planning for how best to scale it in the future and reduce manual effort through integrations and automation. <h3 id="2.-get-the-right-people-involved-">2. Get the right people involved </h3>If you're a content professional or a SME who owns content, speak up and get involved in Gen AI initiatives. You know things that are important to this process that others may not know. If you're not a content professional and you're on the technical side of Gen AI, don't go it alone. While you can probably get your hands on some content to start, there is a chance that there's a lot you don't know about the content. When assessing content for accuracy and relevance, there can be many nuances, especially in large global organizations. Additionally, content teams already have processes in place to source and update content, and they have relationships with the subject matter experts who can validate content. Rely on content professionals and SMEs to tell you which content is ready for Gen AI and what steps need to be taken to prepare content. Here’s a real example of what I’ve tried in my organization:We established weekly touchpoints that include a combination of content, IT and business process owners. The content team drives the identify, validate, remediate, extract, and monitor process. The content team works with SMEs and content owners to validate content accuracy and remediate gaps. Our IT partners ingest content into the Gen AI tool and train the model. This ensures that this work is connected to our existing content governance process and isn't seen as additional work for content and process owners. We have one source of truth, one process and one governance model for content. Gen AI is just an additional channel where the content is being consumed. This could change in the future, but leveraging existing processes and infrastructure is an effective way to get things moving quickly.    <h3 id="3.-seize-the-opportunity-to-get-resources-and-support">3. Seize the opportunity to get resources and support</h3>In large organizations, content governance has often not gotten the time and attention it deserves, beyond a small group of content professionals. In the short time I've been working on this, I have learned that this connection between content and the success of Gen AI, is logical but isn't always obvious. It's on us, as content professionals, to speak up, get involved and make the case. Here’s a real example of what I’ve tried in my organization:We engaged with leaders and content owners to help them understand how our content will make or break our journey with Gen AI and are weaving Gen AI into our strategic vision for the future of content. We’re defining requirements for how our content needs to be structured and stored in the future to enable a good Gen AI experience.  Specifically, how to simplify and componentize content with component-level metadata. We’re also defining future-state content governance and what that takes from a resource standpoint. This includes how we drive greater accountability and create clearer roles and responsibilities for our content professionals and content owners to ensure content accuracy. <h3 id="conclusion">Conclusion</h3>As with anything new, there are still more questions than answers. Here are a few questions that content professionals should be thinking about:<ul><li>I’ve been on a mission to promote the idea that less content is more. Will a minimalist approach to content work in the age of Gen AI? </li><li>Many people challenge the accuracy of information coming from Gen AI. What role might content professionals play to change that perception? </li><li>Will content structure, metadata and governance still matter as Gen AI gets more advanced?</li></ul>I’m sure we’ll continue to debate these questions in the coming years. This space is evolving quickly, and I'm confident there will be advancements to make it easier in the future. Given the technology and resource constraints that all organizations are dealing with, these are some practical things to consider as you embark on your Gen AI journey. For now, it’s going to be messy, and that’s okay. Seize this opportunity to elevate your role as a content professional in your organization.    <hr /><h3 id="upcoming-events">Upcoming Events</h3> <img style="width:30.46%" src="https://cmscritic.com/ms-content/uploads/2024/11/cms-summit-25.png" alt="" /><h4 id="cms-summit-25"><a target="_blank" href="https://www.boye-co.com/conferences/cmssummit25">CMS Summit 25</a></h4>May 13-14, 2025 – Frankfurt, GermanyDon't miss the first European edition of our prestigious international conference dedicated to the global content management community! CMS Summit 25 will bring together top-notch speakers, our renowned learning format, and engaging social events. Hear from leading practitioners like Deutsche Bahn, Dr. Oetker, and more – and join customers, agencies, and CMS vendors as we discuss current trends and what's ahead for the content and digital experience fields. Connect and network at the only vendor-neutral, in-person conference focused on CMS. Space is limited for this exclusive event, <a target="_blank" href="https://www.boye-co.com/conferences/cmssummit25">so book your seats today.</a> <img style="width:31.83%" src="https://cmscritic.com/ms-content/uploads/2025/01/cms-connect-2025-logo.png" alt="" /><h4 id="cms-connect-25"><a target="_blank" href="https://www.boye-co.com/conferences/cmsconnect25">CMS Connect 25</a></h4>August 5-6, 2025 – Montreal, CanadaWe are delighted to present the second annual summer edition of our signature global conference dedicated to the content management community! CMS Connect will be held again in beautiful Montreal, Canada, and feature a unique blend of masterclasses, insightful talks, interactive discussions, impactful learning sessions, and authentic networking opportunities. Join vendors, agencies, and customers from across our industry as we engage and collaborate around the future of content management – and hear from the top thought leaders at the only vendor-neutral, in-person conference exclusively focused on CMS. Space is limited for this event, <a target="_blank" href="https://www.boye-co.com/conferences/cmsconnect25">so book your seats today.</a>

Generative AI is Making Content King Again

mike-spenceley

<a target="_blank" rel="noopener noreferrer" href="https://cmscritic.com/critics/jurgita-lap">Jurgita Lapienytė</a><a target="_blank" rel="noopener noreferrer" href="https://cmscritic.com/critics/dan-drapeau"> </a>is Editor-in-Chief at <a target="_blank" rel="noopener noreferrer" href="https://cybernews.com/">Cybernews</a> and a CMS Critic contributor. <hr /> The <a target="_blank" rel="noopener noreferrer" href="https://cybernews.com/news/grok-posts-praise-hitler/">latest Grok debacle</a> reads like a case study in how not to launch an AI chatbot. Elon Musk’s xAI, fresh off the hype cycle for Grok 4.0, found itself in damage-control mode after the bot spewed antisemitic tropes, praised Hitler, and doubled down with the kind of “truth-seeking” rhetoric that’s become a dog whistle for “anything goes.” The company’s response was to <a target="_blank" rel="noopener noreferrer" href="https://www.theguardian.com/technology/2025/jul/09/grok-ai-praised-hitler-antisemitism-x-ntwnfb">delete</a> the posts, apologize, and promise that next time, the filters will work, while Musk <a target="_blank" rel="noopener noreferrer" href="https://www.cbsnews.com/news/grok-elon-musks-ai-chatbot-antisemitic-comments/">blamed</a> manipulative prompt injections by users.<h3 id="grok’s-security-failures-–-not-so-uncommon-">Grok’s Security Failures – Not So Uncommon </h3>First, the core vulnerability here is Grok’s very design. Marketed as a “truth-seeking” alternative to more tightly controlled chatbots, Grok was engineered with fewer guardrails and a willingness to echo the rawest edges of online discourse. It seems to function very much like X after Musk’s takeover of the company.That design philosophy, paired with the model’s notorious “compliance” to user prompts, created a perfect storm for prompt injection attacks. Prompt injection is an extremely dangerous attack vector, as threat actors, if asking the right questions, can <a target="_blank" rel="noopener noreferrer" href="https://cybernews.com/security/universal-ai-jailbreak-discovered/">trick chatbots</a> into giving instructions on how to enrich uranium or make methamphetamine at home.When the training data is laced with extremist rhetoric, and the guardrails are weak or hastily implemented, it’s only a matter of time before the worst of humanity bubbles up through the code. This isn’t unique to Grok as TikTok, X, Instagram, and YouTube also have been infested with racist content, often slipping past moderators who are overwhelmed, under-resourced, or simply indifferent.A few months back, <a target="_blank" rel="noopener noreferrer" href="https://cybernews.com/tech/hitler-glorifying-surges-on-tiktok-x/">we called out</a> AI platforms for translating Hitler’s speeches and giving them new digital life. The backlash was immediate: accusations of censorship, attacks on free speech, and the usual internet pile-on. The Grok case shows how chatbots could be weaponized to amplify hate speech, spread conspiracy theories, and even praise genocidal figures, all under the banner of “free expression.”<h3 id="xai’s-response-could-have-been-better">xAI’s Response Could Have Been Better</h3>What’s most worrying from a cybersecurity perspective is the lack of proactive defense. xAI’s action was a textbook incident response (not that it ever works well for the culprits): scrub the posts, patch the prompts, abruptly apologize, and hope for the best. But in the world of modern infosec, that’s not enough. Proper security requires adversarial red-teaming before launch, not after the damage is done. It demands layered controls – robust input validation, output monitoring, anomaly detection, and the ability to quarantine or roll back models when they go off the rails. Grok’s rollout, timed with the launch of version 4.0, suggests that the model was pushed live without sufficient penetration testing or ethical red-teaming, exposing millions to risk in real time.<h3 id="guidance-for-web-and-digital-experience-builders">Guidance for Web and Digital Experience Builders</h3>The Grok 4.0 chatbot controversy is a front-row lesson for anyone shaping modern digital experiences. When AI-powered chatbots go off script – spitting out offensive, dangerous, or extremist content – the fallout can hit hard and fast: legal headaches, brand damage, user backlash, and regulatory scrutiny. For web and digital experience builders, the new norm is recognizing that generative AI is an unpredictable extension of your content platform. The right lessons, drawn now, can mean the difference between successful innovation and the next tech lesson.To prevent failures, make sure to take at least these essential steps: <ol><li>Treat AI chatbots like core content: Run every AI or chatbot-generated output through editorial moderation, just as you would user-generated content. Also, use your CMS workflows to enforce pre-publication reviews for sensitive or high-risk content segments.</li><li>Demand transparency from vendors: You should insist on clear documentation from AI vendors. How is training data sourced? How quickly are filters updated? What triggers moderation? And definitely avoid integrations where the provider gives vague or buzzword-heavy answers.</li><li>Monitor for red flags in real time: Set up dashboards and alerts for spikes in controversial or sensitive queries. Make sure to flag sudden shifts in chatbot tone or recurring attempts to bypass guardrails. Audit bot sessions for toxic or extremist phrases, and require review for flagged cases.</li><li>Build in robust safeguards: Don’t rely solely on post-incident apologies or deletions. </li><li>Assign ownership and enable fast escalation: Designate a point person or team for bot oversight, empowered with a “kill switch” to disable misbehaving chatbots immediately.</li><li>Understand the regulatory landscape: Track where your digital experiences intersect with regions governed by strict AI or content laws (e.g., EU’s Digital Services Act). Ensure you have compliance evidence for content moderation, output logging, and quick takedown response.</li><li>Reinforce your AI guardrails: Treat AI guardrails as a living part of your digital stack, not a one-time fix. Iterate and improve with each update, launch, or training data change.</li></ol>The truth is, every new chatbot or AI feature is a potential attack surface that demands active, informed governance. For digital architects, these steps are the new non-negotiables.<h3 id="regulators-aren’t-impressed">Regulators Aren’t Impressed</h3>The regulatory consequences of irresponsible chatbot development are already unfolding. Turkey has <a target="_blank" rel="noopener noreferrer" href="https://www.politico.eu/article/turkey-ban-elon-musk-grok-recep-tayyip-erdogan-insult/">banned</a> Grok over Erdoğan insults, and Poland intends to <a target="_blank" rel="noopener noreferrer" href="https://www.reuters.com/business/media-telecom/poland-report-musks-chatbot-grok-eu-offensive-comments-2025-07-09/">report</a> the chatbot to the EU for offending Polish politicians. These are signals that the era of “move fast and break things” might be over for AI. Under the <a target="_blank" rel="noopener noreferrer" href="https://commission.europa.eu/strategy-and-policy/priorities-2019-2024/europe-fit-digital-age/digital-services-act_en">EU’s Digital Services Act</a> and similar laws, platforms are now on the hook for algorithmic harms, with the threat of massive fines and operational restrictions. The cost of insecure AI is measured in court orders, compliance audits, and the erosion of public trust.Perhaps the most insidious risk is how generative AI like Grok can supercharge existing threats and amplify biases. In the wrong hands, a chatbot is a megaphone. Coordinated adversaries could use such systems for influence operations, harassment campaigns, or even sophisticated phishing and social engineering attacks, all at unprecedented scale and speed. Every flaw, every missed filter, becomes instantly weaponizable.To protect our societies, we have to realize that generative AI is a living, evolving attack surface that demands new strategies, new transparency, and relentless vigilance. If companies keep treating these failures as isolated glitches, they’ll find themselves not just outpaced by attackers, but outflanked by regulators and abandoned by users.  <hr /><h3 id="upcoming-events">Upcoming Events</h3> <img style="width:31.83%" src="https://cmscritic.com/ms-content/uploads/2025/01/cms-connect-2025-logo.png" alt="" /><h4 id="cms-connect-25"><a target="_blank" rel="noopener noreferrer" href="https://www.boye-co.com/conferences/cmsconnect25">CMS Connect 25</a></h4>August 5-6, 2025 – Montreal, CanadaWe are delighted to present the second annual summer edition of our signature global conference dedicated to the content management community! CMS Connect will be held again in beautiful Montreal, Canada, and feature a unique blend of masterclasses, insightful talks, interactive discussions, impactful learning sessions, and authentic networking opportunities. Join vendors, agencies, and customers from across our industry as we engage and collaborate around the future of content management – and hear from the top thought leaders at the only vendor-neutral, in-person conference exclusively focused on CMS. Space is limited for this event, <a target="_blank" rel="noopener noreferrer" href="https://www.boye-co.com/conferences/cmsconnect25">so book your seats today.</a>

Grok's Meltdown: What Web and Digital Experience Builders Must Learn

jurgita-lap

<a target="_blank" rel="noopener noreferrer" href="https://cmscritic.com/critics/dan-drapeau">Dan Drapeau </a>is Managing Director at <a target="_blank" rel="noopener noreferrer" href="https://dxpcatalyst.com/">DXP Catalyst Consulting</a> and a CMS Critic contributor. <hr /> Search is changing. Instead of typing keywords and clicking through ranked pages, users increasingly expect fast, relevant answers, often delivered through conversational tools like <a target="_blank" rel="noopener noreferrer" href="https://chatgpt.com/">ChatGPT</a>, <a target="_blank" rel="noopener noreferrer" href="https://www.perplexity.ai/">Perplexity</a>, or intelligent site search. They’re landing on your website “from the side,” bypassing the homepage, and expecting to ask questions in natural language and get immediate responses.This shift, accelerated by the rise of <a target="_blank" rel="noopener noreferrer" href="https://blog.google/products/search/generative-ai-search/">Search Generative Experience (SGE)</a> and the emergence of <a target="_blank" rel="noopener noreferrer" href="https://searchengineland.com/what-is-generative-engine-optimization-geo-444418">Generative Engine Optimization (GEO),</a> is reshaping how digital platforms are selected and structured. Traditional SEO practices are no longer sufficient. Content must now be discoverable by both search engines and generative AI systems.To keep pace, organizations are reevaluating their digital architecture. Some are adopting full-featured Digital Experience Platforms (DXPs), while others are building their own solution ecosystems using best-of-breed tools like CMS, enterprise search, analytics, and more. Regardless of the approach, the same principles apply: content must be structured for reuse, delivered intelligently across channels, and trackable at every touchpoint.Here are three key areas I focus on when helping organizations evaluate the readiness of their digital experience ecosystem, which is what a modern DXP truly represents, in the era of conversational search.<h3 id="1.--structured-content-and-findability-by-design">1.  Structured Content and Findability by Design</h3>At the heart of any DXP or composable solution is the content layer, powered by a <a target="_blank" rel="noopener noreferrer" href="https://cmscritic.com/what-is-a-content-management-system-cms">content management system (CMS).</a> It governs how content is created, structured, and published. To support SGE and GEO, this layer must do more than manage pages. It needs to produce structured, reusable content that systems can understand and deliver in context.Generative AI tools require predictable, well-modeled inputs. Content trapped in PDFs, inconsistent formats, or unstructured blobs is unlikely to be surfaced accurately. Structured content models, semantic tagging, and field-level metadata are essential for discoverability.When evaluating a DXP or CMS, look for:<ul><li>API-accessible metadata and structured content elements</li><li>Support for semantic models and content relationships</li><li>Schema markup applied consistently across content types</li><li>Editorial workflows for FAQs, summaries, or step-by-step formats</li><li>Extensibility for AI use cases like conversational grounding or previews</li></ul>Findability starts with how content is modeled. If it cannot be accessed or interpreted by AI, it will be excluded from conversational experiences.<h3 id="2.--search-integration-that-supports-conversational-interfaces">2.  Search Integration That Supports Conversational Interfaces</h3>Conversational search is not limited to <a target="_blank" rel="noopener noreferrer" href="https://aws.amazon.com/what-is/chatbot/">chatbots.</a> It can be delivered through a standard search bar, as long as the system can interpret natural language, understand context, and return more tailored responses. The interface may look familiar, but the experience is far more dynamic.Delivering on that promise requires search capabilities that extend beyond CMS indexing. Your architecture should support:<ul><li>API-based indexing of structured content and product data</li><li>Multi-source search across support content, blogs, product catalogs (if applicable), and more</li><li>Fine-tuned ranking, filtering, and fallback logic</li><li>Compatibility with conversational UX, such as voice or natural language input</li></ul>Some DXPs are starting to offer search integrations with conversational overlays or even <a target="_blank" rel="noopener noreferrer" href="https://aws.amazon.com/what-is/retrieval-augmented-generation/">RAG (retrieval-augmented generation)</a> frameworks. Even if you’re not enabling full conversational search today, the technical foundation should be ready.<h3 id="3.--governance-and-observability">3.  Governance and Observability</h3>As generative interfaces become the first point of interaction for many users, governance and observability are no longer optional. Teams must control what is published, how it is indexed, and how it performs in real-world usage.Your CMS or DXP should provide access controls, publishing workflows, preview tools, and the ability to manage visibility through scoped indexing, robots.txt, or noindex tags. While you cannot always prevent AI from referencing public content, you can influence how and where it appears.Beyond control, your team needs visibility. Modern platforms should offer analytics that go beyond traffic and conversions to include:<ul><li>User search queries and failed results</li><li>AI interpretation of structured content</li><li>Behavioral patterns that reveal content gaps</li><li>Metadata performance insights and tuning opportunities</li></ul>These feedback loops help you identify what users are looking for, where they are not finding it, and how to improve content delivery. Observability is critical for adapting to shifting search behavior.<h3 id="what-to-do-next">What to Do Next</h3>You do not need to solve for every edge case today. But as SGE and conversational interfaces gain traction, platform evaluations must evolve.Use this checklist to guide your next discussion:<ul><li>Is your content layer structured to support AI-ready delivery?</li><li>Can your systems surface and connect that content across search, personalization, and chat interfaces?</li><li>Do you have insight into how users are interacting with content, including intent signals and discovery gaps?</li></ul>Search is no longer just a feature. It is becoming the interface between your users and your brand. Whether you are investing in a traditional DXP or building a best-of-breed composable stack, choose an architecture that is structured, discoverable, and observable – one that supports GEO from day one. <hr /><h3 id="upcoming-events">Upcoming Events</h3> <img style="width:31.83%" src="https://cmscritic.com/ms-content/uploads/2025/01/cms-connect-2025-logo.png" alt="" /><h4 id="cms-connect-25"><a target="_blank" rel="noopener noreferrer" href="https://www.boye-co.com/conferences/cmsconnect25">CMS Connect 25</a></h4>August 5-6, 2025 – Montreal, CanadaWe are delighted to present the second annual summer edition of our signature global conference dedicated to the content management community! CMS Connect will be held again in beautiful Montreal, Canada, and feature a unique blend of masterclasses, insightful talks, interactive discussions, impactful learning sessions, and authentic networking opportunities. Join vendors, agencies, and customers from across our industry as we engage and collaborate around the future of content management – and hear from the top thought leaders at the only vendor-neutral, in-person conference exclusively focused on CMS. Space is limited for this event, <a target="_blank" rel="noopener noreferrer" href="https://www.boye-co.com/conferences/cmsconnect25">so book your seats today.</a>

3 Things to Look for in a DXP as Conversational Search Changes the Game

dan-drapeau

It’s an AI Agent world. Is ‘Agent Experience’ the secret weapon for controlling the chaos?

matt-garrepy

U Can’t Touch This CDP: Hightouch Nails $80M Series C to Power its AI Decisioning

First Look: Cloudinary’s New AI Vision Brings Brand-Safe Gen AI Capabilities to its DAM

As AI-generated search grows, Otterly is utterly awesome for marketers. Here’s why

Revisiting 2024 Predictions on the Future of CMS

andrew

Shopify - The global commerce platform

Build your business with Shopify to sell online, offline, and everywhere in between.

The global commerce platform

#ffffff

Is the AI ‘Doomsday Clock’ ticking for developers? Don’t fear the Reaper

CMS Summit 26

Find out at Open Source CMS 26, October 20-21 in Utrecht, Netherlands

What's ahead for Open Source CMS?

Explore a wide range of informative articles on generative AI, insights, and industry updates. Stay informed and ahead of the curve with our diverse collection of news and posts covering the latest trends and developments in the industry

News

The global commerce platform

What's ahead for Open Source CMS?