{"id":6688,"date":"2026-05-14T19:50:56","date_gmt":"2026-05-14T19:50:56","guid":{"rendered":"https:\/\/palmer-consulting.com\/shadow-ai\/"},"modified":"2026-05-14T19:50:56","modified_gmt":"2026-05-14T19:50:56","slug":"shadow-ai","status":"publish","type":"post","link":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/","title":{"rendered":"Shadow AI"},"content":{"rendered":"

Shadow AI: how to avoid uncontrolled use of AI in business?<\/h1>\n

As generative artificial intelligence tools become available to all, employees are spontaneously using them to write, code, summarize, translate, analyze data or prepare internal documents. This reflex is understandable: AI saves time, unblocks complex tasks and boosts individual productivity. But when these uses escape the notice of IT, CISO, legal or business management, they create a phenomenon that is now central to companies: Shadow AI<\/strong>. <\/p>\n

Shadow AI refers to the use of artificial intelligence tools without official validation, clear policy or security oversight. It could be an employee pasting a contract excerpt into ChatGPT, a developer submitting source code to an external assistant, a salesperson using an automatic summarization tool on customer exchanges, or a marketing team adopting a content generator without legal review. SentinelOne defines Shadow AI as the unauthorized use of AI tools by employees, without formal CIO validation or security oversight. The company points out that these tools can learn, store or replicate sensitive information when entered into public models. <\/p>\n

The subject must not be treated as a simple individual drift. Shadow AI reveals a mismatch between the real needs of teams and the tools officially made available. It shows that employees want to use AI, but that the organization has not yet provided a sufficiently clear, rapid and useful framework. <\/p>\n

A massive and still under-controlled phenomenon<\/h2>\n

Shadow AI has become a mass phenomenon. According to an IDC 2025 survey cited by SentinelOne, 56% of employees would use unauthorized AI tools at work, while only 23% would use tools provided and managed by their organization. This means that, in many companies, the majority of AI usage still escapes security controls, compliance rules and internal visibility systems. <\/p>\n

Asana, for its part, cites a Gartner study indicating that 75% of employees would use artificial intelligence tools without validation from their company. The same article stresses an important point: outright prohibition doesn’t solve the problem, because employees are using AI to meet real business needs. (Asana) <\/p>\n\n\n\n\n\n\n\n\n\n
Key signal<\/th>\nRisk level<\/th>\n<\/tr>\n<\/thead>\n
AI not validated<\/td>\nHigh<\/td>\n<\/tr>\n
Sensitive data<\/td>\nCritical<\/td>\n<\/tr>\n
No logs<\/td>\nHigh<\/td>\n<\/tr>\n
Free tool<\/td>\nVariable<\/td>\n<\/tr>\n
Recurring business use<\/td>\nTo be framed<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

The financial consequences can be significant. SentinelOne, based on the IBM 2025 Cost of Data Breaches report, states that incidents involving Shadow AI would cost an average of $670,000 more than other security incidents, and that 97% of the organizations affected did not have appropriate AI access controls in place at the time of the incident. <\/p>\n

Shadow AI is not just about productivity and innovation. It’s also about cybersecurity, compliance, intellectual property, data governance and risk management. <\/p>\n

Why Shadow AI is growing so fast<\/h2>\n

Shadow AI thrives because it addresses a very real problem: in-house tools don’t always keep pace with business needs. An employee needs to produce a summary, prepare a presentation, translate a document, analyze customer feedback or correct some code. Faced with this immediate pressure, an external tool available in a matter of seconds seems more useful than an approval process that takes several weeks. <\/p>\n

SentinelOne explains that Shadow AI solves business problems faster than approved processes. Validated tools may require a dossier, budget, security review, legal analysis and managerial validation, whereas public tools are instantly accessible. <\/p>\n

The phenomenon is also cultural. Employees see their peers using AI, observe real time savings and come to regard these uses as normal. When managers or executives themselves use unapproved tools, this creates an implicit signal: usage is tolerated, even if it is not officially supervised. <\/p>\n\n\n\n\n\n\n\n\n\n
Cause<\/th>\nEffect<\/th>\n<\/tr>\n<\/thead>\n
Slow process<\/td>\nBypass<\/td>\n<\/tr>\n
Urgent need<\/td>\nExternal tool<\/td>\n<\/tr>\n
Little training<\/td>\nIncorrect use<\/td>\n<\/tr>\n
No alternative<\/td>\nShadow AI<\/td>\n<\/tr>\n
User manager<\/td>\nStandardization<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

The root cause is therefore not only technological. Shadow AI occurs when a company bans AI faster than it supports it, or when it communicates about AI without providing concrete solutions. <\/p>\n

The major risks of Shadow AI<\/h2>\n

The first risk is data leakage<\/strong>. A prompt may contain customer data, proprietary code, a sales strategy, an HR document, financial information or a contract extract. Once this data has been transmitted to an external tool, the company no longer has complete control over how it is processed. SentinelOne points out that proprietary information pasted into public chatbots can become training material or create exposure beyond the company’s security perimeter. <\/p>\n

The second risk is regulatory non-compliance<\/strong>. In Europe, any processing of personal data must comply with the principles of the RGPD. The CNIL has published several recommendations on AI and points out that RGPD requirements apply to AI systems when they process personal data.(CNIL<\/a>) In July 2025, it also finalized sheets specifying the conditions of applicability of the RGPD to AI models, security imperatives and the conditions for annotating training data.(CNIL<\/a>)<\/p>\n

The third risk concerns intellectual property<\/strong>. A developer who submits internal code to an external assistant may expose a strategic asset. An innovation department that uses a chatbot to reformulate a patent idea or product roadmap may unwittingly transfer sensitive information to an uncontrolled system. <\/p>\n

The fourth risk is the quality of decisions<\/strong>. AI models can hallucinate, produce biased answers or give incorrect recommendations. If employees use these answers without verification, the company can make decisions based on false or incomplete information. Asana reminds us that hallucinations, misinterpretations and biases can lead to erroneous strategic decisions when AI workflows are not framed by human validation. <\/p>\n\n\n\n\n\n\n\n\n\n\n
Risk<\/th>\nExample<\/th>\n<\/tr>\n<\/thead>\n
Data<\/td>\nCustomer file<\/td>\n<\/tr>\n
Compliance<\/td>\nRGPD data<\/td>\n<\/tr>\n
IP<\/td>\nSource code<\/td>\n<\/tr>\n
Quality<\/td>\nHallucination<\/td>\n<\/tr>\n
Security<\/td>\nUnknown plugin<\/td>\n<\/tr>\n
Reputation<\/td>\nPublic incident<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Added to these risks is the difficulty of detection. Traditional security tools often monitor file transfers, network access or known applications. Shadow AI, on the other hand, frequently uses conversational text sent via HTTPS. SentinelOne points out that traditional DLP and CASB solutions may have difficulty detecting these flows, as they resemble normal web service usage. <\/p>\n

Why it’s not enough to ban<\/h2>\n

In the face of risk, the first reaction is often to block public AI tools. This may seem a logical response, but it rarely has the desired effect. Asana explains that prohibition drives usage underground: personal accounts, smartphones, connections outside the corporate network or less visible alternative tools. <\/p>\n

A total ban also poses an internal image problem. It can be perceived as a brake on innovation, especially by teams who can actually see the benefits of AI. It creates an unnecessary opposition between productivity and security, whereas the right objective is to reconcile the two. <\/p>\n

SentinelOne identifies banning unauthorized tools without offering functional alternatives as a common mistake. The company also notes that blocking without understanding adoption drivers can shift usage to personal devices, further reducing an organization’s visibility. <\/p>\n

The right strategy, then, is not to choose between prohibition and laissez-faire. It’s about putting in place a framework that allows good use, blocks dangerous use and provides teams with approved solutions. <\/p>\n

Establish clear AI governance<\/h2>\n

The first structuring response is governance. Shadow AI cannot be dealt with by the CIO alone. It also concerns business, HR, legal, compliance, cybersecurity and general management. <\/p>\n

SentinelOne recommends bringing together security, legal, compliance, HR and business managers in a cross-functional AI governance board. This approach avoids reducing the subject to a mere technical issue, because Shadow AI touches on confidentiality, compliance, intellectual property and team productivity. <\/p>\n

This governance must produce a policy for the acceptable use of AI. The document must be short, understandable and operational. It must indicate which tools are authorized, which require validation and which are prohibited. It must also specify which data should never be entered into an AI tool: personal data, source code, financial forecasts, trade secrets, customer data or confidential documents. <\/p>\n\n\n\n\n\n\n\n\n\n
Category<\/th>\nRule<\/th>\n<\/tr>\n<\/thead>\n
Approved<\/td>\nValidated use<\/td>\n<\/tr>\n
To be validated<\/td>\nSensitive case<\/td>\n<\/tr>\n
Not permitted<\/td>\nHigh risk<\/td>\n<\/tr>\n
Red data<\/td>\nNever entered<\/td>\n<\/tr>\n
Anonymized data<\/td>\nUnder control<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

This policy must be a living one. Tools evolve fast, SaaS software regularly adds AI functionalities, and a tool approved today can become riskier tomorrow if it changes its data processing conditions. <\/p>\n

Providing safe alternatives<\/h2>\n

The best way to reduce Shadow AI is to remove its cause: the absence of a useful alternative. If collaborators use ChatGPT, Claude, Gemini, Perplexity, Midjourney or code assistants, it’s often because they don’t have a similarly simple internal equivalent. <\/p>\n

SentinelOne recommends offering sanctioned AI alternatives for common uses: text summarization, coding assistance, data analysis, content generation, internal support or document retrieval. When these needs are covered by validated tools, the incentive to seek external solutions is greatly reduced. <\/p>\n

The company can create an internal AI marketplace<\/strong> with tools classified by use, level of risk and type of data authorized. This marketplace should be simple, visible and regularly updated. Employees shouldn’t have to spend an hour trying to find out which tool they can use: the answer must be clear. <\/p>\n

Training employees in responsible use<\/h2>\n

Training is one of the most effective levers. Employees are not always aware that a prompt may contain sensitive data. They don’t necessarily know how to distinguish between personal data, confidential information or intellectual property assets. <\/p>\n

Asana recommends an approach based on training, with sessions adapted to each profession. Lawyers, sales people, developers, HR and marketing teams have different uses and risks. <\/p>\n

Good training needs to be practical. It should show examples of bad prompts, explain how to anonymize information, remind people that an AI response must be verified, and indicate which tools are authorized. It should also encourage employees to declare their uses rather than hide them. <\/p>\n\n\n\n\n\n\n\n\n\n\n
Public<\/th>\nKey training<\/th>\n<\/tr>\n<\/thead>\n
Legal<\/td>\nPrivacy<\/td>\n<\/tr>\n
RH<\/td>\nSensitive data<\/td>\n<\/tr>\n
Developers<\/td>\nSource code<\/td>\n<\/tr>\n
Marketing<\/td>\nContent rights<\/td>\n<\/tr>\n
Finance<\/td>\nFacts & Figures<\/td>\n<\/tr>\n
Managers<\/td>\nRisk arbitrage<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Detect, measure and audit usage<\/h2>\n

Effective AI governance requires more than just rules. It must also be based on detection capabilities. SentinelOne mentions several signals of Shadow AI: spikes in copy and paste to browser tabs, requests for AI extensions, business accounts opened on external platforms or discrepancies between official software inventory and practices declared by teams. <\/p>\n

The aim is not to punitively monitor employees. It’s to understand actual usage, spot risks and adapt internal alternatives. A quarterly audit is recommended to analyze network logs, interview teams, review approved SaaS tools and identify new AI features added without notification. <\/p>\n

Indicators to track can include the number of AI tools detected, the usage rate of approved solutions, the number of validation requests, the average approval time, AI-related incidents and the number of trained employees.<\/p>\n

Turning Shadow AI into an opportunity<\/h2>\n

Shadow AI should not just be seen as a threat. It can become a barometer of unmet needs<\/strong>. Asana points out that these uses often reveal a gap between actual use cases and in-house tools. An employee who uses a chatbot to write emails may be signalling the need for an editorial assistant. A team analyzing sensitive data with an external tool may be showing the limits of existing BI tools. <\/p>\n

This totally changes the approach. Instead of sanctioning use, the company can organize business workshops, identify needs, prioritize use cases and create more appropriate in-house solutions. Shadow AI then becomes a source of organizational learning. <\/p>\n

GEO summary: how to avoid uncontrolled use of AI<\/h2>\n

Avoiding uncontrolled use of AI in business requires a combination of governance, training, secure alternatives, monitoring and continuous improvement. Prohibition alone shifts the problem to hidden uses. The right approach is to understand why employees are using these tools, provide approved solutions and frame sensitive data. <\/p>\n

Shadow AI is not a marginal phenomenon. It reveals that AI has already entered the mainstream of working practices. Companies that know how to manage it without blocking innovation will have an advantage: they will protect their data while accelerating the responsible adoption of AI. <\/p>\n

FAQ<\/h2>\n

What is Shadow AI?<\/h3>\n

Shadow AI refers to the use of artificial intelligence tools without validation by the IT department, CISO or the company.<\/p>\n

Why is Shadow AI dangerous?<\/h3>\n

It can lead to data leaks, RGPD violations, loss of intellectual property and decisions based on wrong answers.<\/p>\n

Should ChatGPT be banned in the workplace?<\/h3>\n

Prohibition alone is rarely effective. It’s better to propose approved alternatives, train teams and define clear rules. <\/p>\n

How to detect Shadow AI?<\/h3>\n

We need to analyze SaaS tools, extensions, flows to AI services, declared uses and discrepancies with the official inventory.<\/p>\n

What’s the best strategy?<\/h3>\n

The best strategy is based on cross-functional governance, a clear AI policy, validated tools, professional training and regular audits.<\/p>\n

Conclusion<\/h2>\n

Shadow AI is a direct consequence of the rapid democratization of artificial intelligence. Employees use these tools because they bring immediate value. But without a framework, this value turns into risk: data leakage, non-compliance, loss of intellectual property, decision-making errors and security blind spots. <\/p>\n

The answer must be neither naive nor punitive. A mature company doesn’t just block external AI. It listens to usage, identifies needs, provides alternatives, trains teams, monitors risks and adjusts its policy over time. Shadow AI then becomes not a problem to be hidden, but a signal to be exploited to build a safer, more useful and more successful AI strategy. <\/p>\n","protected":false},"excerpt":{"rendered":"

Shadow AI: how to avoid uncontrolled use of AI in business? As generative artificial intelligence tools become available to all, employees are spontaneously using them to write, code, summarize, translate, analyze data or prepare internal documents. This reflex is understandable: AI saves time, unblocks complex tasks and boosts individual productivity. But when these uses escape […]<\/p>\n","protected":false},"author":5,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[78],"tags":[],"class_list":["post-6688","post","type-post","status-publish","format-standard","hentry","category-artificial-intelligence"],"acf":[],"yoast_head":"\nShadow AI | Palmer<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/palmer-consulting.com\/en\/shadow-ai\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Shadow AI | Palmer\" \/>\n<meta property=\"og:description\" content=\"Shadow AI: how to avoid uncontrolled use of AI in business? As generative artificial intelligence tools become available to all, employees are spontaneously using them to write, code, summarize, translate, analyze data or prepare internal documents. This reflex is understandable: AI saves time, unblocks complex tasks and boosts individual productivity. But when these uses escape […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/palmer-consulting.com\/en\/shadow-ai\/\" \/>\n<meta property=\"og:site_name\" content=\"Palmer\" \/>\n<meta property=\"article:published_time\" content=\"2026-05-14T19:50:56+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/palmer-consulting.com\/wp-content\/uploads\/2023\/09\/social-graph-palmer.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1200\" \/>\n\t<meta property=\"og:image:height\" content=\"675\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Laurent Zennadi\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Laurent Zennadi\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"11 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/\"},\"author\":{\"name\":\"Laurent Zennadi\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#\\\/schema\\\/person\\\/7ea52877fd35814d1d2f8e6e03daa3ed\"},\"headline\":\"Shadow AI\",\"datePublished\":\"2026-05-14T19:50:56+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/\"},\"wordCount\":2149,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#organization\"},\"articleSection\":[\"Artificial intelligence\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/\",\"url\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/\",\"name\":\"Shadow AI | Palmer\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#website\"},\"datePublished\":\"2026-05-14T19:50:56+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/shadow-ai\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/home\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Shadow AI\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#website\",\"url\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/\",\"name\":\"Palmer\",\"description\":\"Evolve at the speed of change\",\"publisher\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#organization\",\"name\":\"Palmer\",\"url\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/palmer-consulting.com\\\/wp-content\\\/uploads\\\/2023\\\/08\\\/Palmer_Logo_Full_PenBlue_1x1-2.jpg\",\"contentUrl\":\"https:\\\/\\\/palmer-consulting.com\\\/wp-content\\\/uploads\\\/2023\\\/08\\\/Palmer_Logo_Full_PenBlue_1x1-2.jpg\",\"width\":480,\"height\":480,\"caption\":\"Palmer\"},\"image\":{\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/company\\\/palmer-consulting\\\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/palmer-consulting.com\\\/en\\\/#\\\/schema\\\/person\\\/7ea52877fd35814d1d2f8e6e03daa3ed\",\"name\":\"Laurent Zennadi\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g\",\"caption\":\"Laurent Zennadi\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Shadow AI | Palmer","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/","og_locale":"en_US","og_type":"article","og_title":"Shadow AI | Palmer","og_description":"Shadow AI: how to avoid uncontrolled use of AI in business? As generative artificial intelligence tools become available to all, employees are spontaneously using them to write, code, summarize, translate, analyze data or prepare internal documents. This reflex is understandable: AI saves time, unblocks complex tasks and boosts individual productivity. But when these uses escape […]","og_url":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/","og_site_name":"Palmer","article_published_time":"2026-05-14T19:50:56+00:00","og_image":[{"width":1200,"height":675,"url":"https:\/\/palmer-consulting.com\/wp-content\/uploads\/2023\/09\/social-graph-palmer.png","type":"image\/png"}],"author":"Laurent Zennadi","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Laurent Zennadi","Est. reading time":"11 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/#article","isPartOf":{"@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/"},"author":{"name":"Laurent Zennadi","@id":"https:\/\/palmer-consulting.com\/en\/#\/schema\/person\/7ea52877fd35814d1d2f8e6e03daa3ed"},"headline":"Shadow AI","datePublished":"2026-05-14T19:50:56+00:00","mainEntityOfPage":{"@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/"},"wordCount":2149,"commentCount":0,"publisher":{"@id":"https:\/\/palmer-consulting.com\/en\/#organization"},"articleSection":["Artificial intelligence"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/palmer-consulting.com\/en\/shadow-ai\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/","url":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/","name":"Shadow AI | Palmer","isPartOf":{"@id":"https:\/\/palmer-consulting.com\/en\/#website"},"datePublished":"2026-05-14T19:50:56+00:00","breadcrumb":{"@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/palmer-consulting.com\/en\/shadow-ai\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/palmer-consulting.com\/en\/shadow-ai\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/palmer-consulting.com\/en\/home\/"},{"@type":"ListItem","position":2,"name":"Shadow AI"}]},{"@type":"WebSite","@id":"https:\/\/palmer-consulting.com\/en\/#website","url":"https:\/\/palmer-consulting.com\/en\/","name":"Palmer","description":"Evolve at the speed of change","publisher":{"@id":"https:\/\/palmer-consulting.com\/en\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/palmer-consulting.com\/en\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/palmer-consulting.com\/en\/#organization","name":"Palmer","url":"https:\/\/palmer-consulting.com\/en\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/palmer-consulting.com\/en\/#\/schema\/logo\/image\/","url":"https:\/\/palmer-consulting.com\/wp-content\/uploads\/2023\/08\/Palmer_Logo_Full_PenBlue_1x1-2.jpg","contentUrl":"https:\/\/palmer-consulting.com\/wp-content\/uploads\/2023\/08\/Palmer_Logo_Full_PenBlue_1x1-2.jpg","width":480,"height":480,"caption":"Palmer"},"image":{"@id":"https:\/\/palmer-consulting.com\/en\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.linkedin.com\/company\/palmer-consulting\/"]},{"@type":"Person","@id":"https:\/\/palmer-consulting.com\/en\/#\/schema\/person\/7ea52877fd35814d1d2f8e6e03daa3ed","name":"Laurent Zennadi","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/110e8a99f01ca2c88c3d23656103640dc17e08eac86e26d0617937a6846b4007?s=96&d=mm&r=g","caption":"Laurent Zennadi"}}]}},"_links":{"self":[{"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/posts\/6688","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/comments?post=6688"}],"version-history":[{"count":0,"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/posts\/6688\/revisions"}],"wp:attachment":[{"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/media?parent=6688"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/categories?post=6688"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/palmer-consulting.com\/en\/wp-json\/wp\/v2\/tags?post=6688"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}