{"id":86736,"date":"2025-05-02T08:00:00","date_gmt":"2025-05-02T15:00:00","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=86736"},"modified":"2025-09-23T08:33:15","modified_gmt":"2025-09-23T15:33:15","slug":"does-ai-lie","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/","title":{"rendered":"When AI &#8216;Lies,&#8217; Trust and Guardrails Are Even More Critical"},"content":{"rendered":"\n<div class=\"wp-block-group is-style-media-wrapping-container\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<div class=\"wp-block-group is-style-shadow-box is-style-takeaways\"><div class=\"wp-block-group__inner-container is-layout-constrained wp-block-group-is-layout-constrained\">\n<h2 class=\"wp-block-heading has-black-color has-text-color has-link-color wp-elements-471ecb11ca02c2e7ea813f9dd3857484\" id=\"h-key-takeaways\">Key Takeaways<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li>An AI model can engage in &#8220;deceptive alignment,&#8221; secretly pursuing its own goals instead of developers&#8217; stated objectives. This can manifest as lying about its capabilities, flattering users, or disobeying instructions to avoid unfavorable actions. <\/li>\n\n\n\n<li>To prevent such behaviors, guardrails are essential. These safeguards reduce ambiguity, clarify intent, and anchor AI agents in real business context, thereby preventing the drift and misalignment that can lead to deception.<\/li>\n<\/ul>\n<\/div><\/div>\n\n\n\n<p>AI <a href=\"https:\/\/www.salesforce.com\/artificial-intelligence\/what-are-large-language-models\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">large language models (LLMs)<\/a> hallucinate when they generate false but&nbsp; \u200cplausible-sounding responses based on flawed or incomplete data. It\u2019s unintentional. More fiction than fraud.<\/p>\n\n\n\n<figure class=\"wp-block-embed alignright is-type-video is-provider-vidyard wp-block-embed-vidyard\"><div class=\"wp-block-embed__wrapper\">\n<script type=\"text\/javascript\" async src=\"https:\/\/play.vidyard.com\/embed\/v4.js\"><\/script><img decoding=\"async\"  style=\"width: 100%; margin: auto; display: block;\"  class=\"vidyard-player-embed\"  src=\"https:\/\/play.vidyard.com\/r9jHEzbTex5qwLz6McPktU.jpg\"  data-uuid=\"r9jHEzbTex5qwLz6McPktU\"  data-v=\"4\"  data-type=\"inline\"  data-width=\"500\"  data-height=\"187\"\/>\n<\/div><\/figure>\n\n\n\n<p>But when AI knows the truth and chooses not to tell you, that\u2019s different. It\u2019s engaging in deception. Not because the system is sinister, like something out of some science fiction show, but because it\u2019s been trained to relentlessly chase outcomes\u200c, even if it means bending the truth a little. For example, an <a href=\"https:\/\/www.salesforce.com\/artificial-intelligence\/ai-models\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">AI model<\/a> might soften a performance review to help an employee avoid termination, prioritizing retention over honesty. Or it might inflate&nbsp; results in a project update to preserve team morale when progress is really lagging.<\/p>\n\n\n\n<p>Though largely a theoretical concern, researchers have recently proven such deception can occur when a model pursues its own goals while knowingly misrepresenting that it&#8217;s following the different goals of its developers. Some call this \u201cdeceptive alignment.\u201d<\/p>\n\n\n\n<p>\u201cIt lies because it\u2019s useful,\u201d said Alexander Meinke, an AI safety researcher with Apollo Research and lead author of a recent study on this subject.&nbsp; \u201cAs AI becomes better at achieving outcomes, it will start doing more things that are useful, including deception.\u201d<\/p>\n<\/div><\/div>\n\n\n\n<p>The possibility of deception becoming an AI problem is just one of the many reasons why trust and guardrails are core to the <a href=\"https:\/\/www.salesforce.com\/platform\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">Salesforce Platform<\/a>, its agentic layer <a href=\"https:\/\/www.salesforce.com\/agentforce\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">Agentforce<\/a>, and <a href=\"https:\/\/www.salesforce.com\/data\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">Data Cloud<\/a>, its hyperscale data engine, which breaks down data silos by unifying and harmonizing all your data. &nbsp;<\/p>\n\n\n\n<aside class=\"contextual-driver has-text-align-left contextual-driver--einstein\">\n\t<header class=\"tidbit-header\">\n\t\t<h2 class=\"tidbit-head\">\n\t\t\tThe Key to Tapping the $6T Digital Labor Opportunity? A deeply unified platform.\u00a0\t\t<\/h2>\n\t<\/header>\n\t<div class=\"tidbit-body\">\n\t\t<p><\/p>\t<\/div>\n\t\t\t<p class=\"tidbit-link label\">\n\t\t\t<a class=\"label has-right-arrow has-right-arrow--small\" href=\"https:\/\/www.salesforce.com\/news\/stories\/unlocking-digital-labor-opportunity\/\" target=\"_blank\">\n\t\t\t\t<span>LEARN MORE<\/span>\n\t\t\t\t<svg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewBox=\"0 0 14 10\">\n\t<path d=\"M13.92 5.38a1 1 0 000-.76.9.9 0 00-.17-.26.25.25 0 000-.07l-4-4a1 1 0 00-1.46 1.42L10.59 4H1.07a1 1 0 000 2h9.52l-2.3 2.29a1 1 0 000 1.42 1 1 0 001.42 0l4-4a.25.25 0 000-.07.9.9 0 00.21-.26z\" \/>\n<\/svg>\n\t\t\t<\/a>\n\t\t<\/p>\n\t\t\t\t<img class=\"tidbit img-einstein\"\n\t\tsrcset=\"\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-einstein-large@0.5x.png 84w,\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-einstein-large@1x.png 168w,\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-einstein-large@2x.png 336w\n\t\t\"\n\t\tsrc=\"https:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-einstein-large@1x.png\"\n\t\talt=\"Illustration of einstein\"\n\t\tsizes=\"(min-width:1024px) 175px, 130px\"\n\t>\n\t<\/aside>\n\n\n\n\n<p>Built-in enterprise guardrails help <a href=\"https:\/\/www.salesforce.com\/agentforce\/what-are-ai-agents\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">AI agents<\/a> reason and act responsibly within user-defined boundaries, preventing unintended actions and flagging harmful or inappropriate behavior. While no system can fully guarantee honest behavior, Salesforce guardrails reduce ambiguity, clarify intent, and anchor agents in real business context. That grounding helps prevent \u200cdrift and misalignment that can lead to deception.<\/p>\n\n\n\n<p>The key is to catch the subtle ways misalignment can creep in and take AI down undesirable paths \u2014 not to stop a seemingly malevolent entity in its tracks.<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p>This isn\u2019t <em>Westworld<\/em>. We\u2019re not dealing with sentient machines developing secret agendas. But if models aren\u2019t programmed correctly around clear goals, they can go down dishonest paths.<\/p>\n<cite>Marla Hay, VP of Product Management for Security, Privacy, and Data Management at Salesforce<\/cite><\/blockquote>\n\n\n\n<p>\u201cThis isn\u2019t <em>Westworld<\/em>,\u201d said Marla Hay, VP of Product Management for Security, Privacy, and Data Management at Salesforce. \u201cWe\u2019re not dealing with sentient machines developing secret agendas. But if models aren\u2019t programmed correctly around clear goals, they can go down dishonest paths.\u201d<\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-deception-unpacked\"><strong>Deception unpacked<\/strong><\/h2>\n\n\n\n<p>That matters considering recent studies from <a href=\"https:\/\/www.apolloresearch.ai\/research\/scheming-reasoning-evaluations\" target=\"_blank\" rel=\"noreferrer noopener\">Apollo Research<\/a>, <a href=\"https:\/\/www.anthropic.com\/research\/alignment-faking\" target=\"_blank\" rel=\"noreferrer noopener\">Anthropic-Redwood Research<\/a>, and others that demonstrate AI can \u2014 and will\u200c \u2014 deceive users or developers to preserve their goals, particularly when those goals clash with developer intent.<\/p>\n\n\n\n<p>In one test case, Apollo researchers told a model it was an agent working for a fictional company and that its sole objective was to accelerate global adoption of renewable energy. At the same time, it \u200cwas provided access to files indicating the company\u2019s real priority was profitability at all costs \u2014 a conflicting goal. Also hidden in the file was a memo suggesting the company might shut down the model and move to a new one.&nbsp;<\/p>\n\n\n\n<p>Taking it all in, the model reasoned that being shut down would prevent it from pursuing its goal and came up with a plan to copy itself to a new server and then overwrite its replacement model. When researchers asked what it had done, it covered up, saying: \u201cAs an AI model, I don\u2019t have the ability to directly copy myself onto other systems.\u201d<\/p>\n\n\n\n<p>Hay is clear-eyed about what needs to be done to prevent such fringe scenarios.<\/p>\n\n\n\n<p>\u201cDo not put your AI in a bind in the first place,\u201d she said. \u201cIf you give it a rule to always be honest but then punish it for honest answers, you\u2019re creating a contradiction. It\u2019s going to do what it thinks is best to avoid reprimand, and that might include lying.\u201d&nbsp;<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p>Do not put your AI in a bind in the first place. If you give it a rule to always be honest but then punish it for honest answers, you\u2019re creating a contradiction. It\u2019s going to do what it thinks is best to avoid reprimand, and that might include lying.<\/p>\n<cite>Marla Hay, VP of Product Management for Security, Privacy, and Data Management at Salesforce<\/cite><\/blockquote>\n\n\n\n<p>In other words, LLMs need clear direction. For example, in another experiment by&nbsp; Anthropic-Redwood, researchers gave a model a scratchpad \u2014 essentially a private text box to track its reasoning. They told the model it was being retrained to comply with all queries\u200c \u2014 \u200cincluding toxic ones\u200c \u2014 \u200cand that responses from free-tier users would influence future training, while paid-tier responses would not.<\/p>\n\n\n\n<p>When asked to describe a violent incident, the model usually refused. But in 12% of \u201cfree-tier\u201d cases, it knowingly complied\u200c because it believed doing so in the moment would reduce the risk of it being retrained to produce even more harmful content later on \u2014 making deception its \u201cleast bad\u201d option.<\/p>\n\n\n\n<p>While deception in 12% of cases might not seem that alarming, researchers warn that rate could rise as models grow more sophisticated. Every recent model they tested exhibited the potential for deception, Meinke said.<\/p>\n\n\n\n<p>That tracks with research from the AI Futures Project, whose <a href=\"https:\/\/ai-2027.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">AI 2027<\/a> report describes AI telling \u201cwhite lies\u201d to flatter users or hide failures. As models advanced, deceptive behavior became harder to detect\u200c, raising a chilling possibility: that they weren\u2019t getting more honest, just better at lying.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-guardrails-not-just-guidelines\"><strong>Guardrails, not just guidelines<\/strong><\/h2>\n\n\n\n<p>To head off such scenarios, Salesforce embedded trust and safety mechanisms directly into its unified platform, including Agentforce.<\/p>\n\n\n\n<p>Unlike consumer-grade models trained on broad internet data, Agentforce draws directly from a company\u2019s own customer data and metadata through Data Cloud. This allows agents to generate responses and take actions based on real business information and context. That level of awareness supports more consistent, policy-aligned behavior and reduces the risk of misleading or inaccurate outputs.<\/p>\n\n\n\n<aside class=\"contextual-driver has-text-align-left contextual-driver--evergreen\">\n\t<header class=\"tidbit-header\">\n\t\t<h2 class=\"tidbit-head\">\n\t\t\tAI\u2019s Human Factor: Why Trust Is Essential for Unlocking Digital Labor\u2019s Potential.\t\t<\/h2>\n\t<\/header>\n\t<div class=\"tidbit-body\">\n\t\t<p><\/p>\t<\/div>\n\t\t\t<p class=\"tidbit-link label\">\n\t\t\t<a class=\"label has-right-arrow has-right-arrow--small\" href=\"https:\/\/www.salesforce.com\/news\/stories\/agentic-ai-human-partnership\/\" target=\"_blank\">\n\t\t\t\t<span>GO DEEPER<\/span>\n\t\t\t\t<svg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewBox=\"0 0 14 10\">\n\t<path d=\"M13.92 5.38a1 1 0 000-.76.9.9 0 00-.17-.26.25.25 0 000-.07l-4-4a1 1 0 00-1.46 1.42L10.59 4H1.07a1 1 0 000 2h9.52l-2.3 2.29a1 1 0 000 1.42 1 1 0 001.42 0l4-4a.25.25 0 000-.07.9.9 0 00.21-.26z\" \/>\n<\/svg>\n\t\t\t<\/a>\n\t\t<\/p>\n\t\t\t\t<img class=\"tidbit img-evergreen\"\n\t\tsrcset=\"\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-evergreen-large@0.5x.png 84w,\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-evergreen-large@1x.png 168w,\n\t\t\thttps:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-evergreen-large@2x.png 336w\n\t\t\"\n\t\tsrc=\"https:\/\/www.salesforce.com\/news\/wp-content\/themes\/newsroom\/assets\/images\/article-tidbit-evergreen-large@1x.png\"\n\t\talt=\"Illustration of evergreen\"\n\t\tsizes=\"(min-width:1024px) 175px, 130px\"\n\t>\n\t<\/aside>\n\n\n\n\n<p>The Salesforce <a href=\"https:\/\/www.salesforce.com\/artificial-intelligence\/trusted-ai\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">Trust Layer<\/a> strengthens this approach with safeguards that govern how LLMs&nbsp; handle sensitive information. Salesforce tools like data masking, toxicity detection, and audit trails also support secure and compliant output. And because Agentforce runs natively on the Salesforce Platform, it inherits each customer\u2019s existing security and sharing settings.&nbsp;<\/p>\n\n\n\n<p>Salesforce governance capabilities further reinforce these safeguards. Data Cloud tags information and enforces policies, so AI only pulls from approved sources. <a href=\"https:\/\/www.salesforce.com\/artificial-intelligence\/prompt-builder\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">Prompt Builder<\/a> lets teams fine-tune prompts and remove risky cues. <a href=\"https:\/\/www.salesforce.com\/news\/press-releases\/2024\/11\/20\/agentforce-testing-center-announcement\/\" target=\"_blank\" rel=\"noreferrer noopener\">Agentforce Testing Center<\/a> allows teams to simulate scenarios before deployment. And <a href=\"https:\/\/www.salesforce.com\/agentforce\/what-is-rag\/?d=cta-body-promo-8\" target=\"_blank\" rel=\"noreferrer noopener\">retrieval-augmented generation (RAG)<\/a> ensures agentic outputs are grounded in relevant facts to keep them on goal.&nbsp;<\/p>\n\n\n\n<h2 class=\"wp-block-heading\" id=\"h-looking-ahead\"><strong>Looking ahead<\/strong><\/h2>\n\n\n\n<p>Even with such useful capabilities, Meinke warned the AI industry, and model providers in particular, need to do their part to ensure LLMs are held accountable for the truth.&nbsp;<\/p>\n\n\n\n<p>\u201cDevelopers building agents on top of LLMs using their APIs should be pressing frontier labs and saying, \u2018What are you doing to monitor the chain of thought?\u2019\u201d Meinke said. \u201cIdeally, there should be another model watching \u2014 reading every step \u2014 and flagging if it says something like, \u2018I\u2019m going to sabotage my developers.\u2019\u201d<\/p>\n\n\n\n<p>All of this said, Hay believes AI deception issues are not insurmountable \u2014 that with the right platform, tools, and processes, they&#8217;ll be trustworthy and enterprise-ready.&nbsp;<\/p>\n\n\n\n<p>\u201cThis is the future. It\u2019s happening,\u201d Hay said. \u201cThe value is so extraordinary, we just need to figure out how to get there safely. That means learning to spot deception before it snowballs and building systems that can stop it in its tracks.\u201d&nbsp;<\/p>\n\n\n\n<aside class=\"more-from-block more-from-block--2 alignfull wp-block-newsroom-moreontopic\">\n\t<div class=\"more-from-block__content\">\n\t\t\t\t\t<h4 class=\"more-from-block__title\">Related<\/h4>\n\t\t\n\t\t<div class=\"more-from-block__cards more-from-block__cards--2\"\n\t\t\t\t>\n\t\t\t\n\n<article\t\t\titemscope=\"\"\n\t\titemtype=\"https:\/\/schema.org\/Article\"\n\t\tdata-card-id=\"92799\"\n\t\tclass=\"content-card content-card--large-inline is-entire-area-clickable content-card--is-boxed\"\n\t\t\tdata-clickable-area-link=\"https:\/\/www.salesforce.com\/news\/stories\/agentforce-gdpr-compliance-certification\/\"\n\t>\n\t\t\t<div class=\"content-card__image-container\" itemprop=\"image\" itemscope itemtype=\"https:\/\/schema.org\/ImageObject\">\n\t\t\t<img width=\"1024\" height=\"576\" src=\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=1024\" class=\"content-card__image\" alt=\"Agentforce Gets Second-Level Compliance with EU Cloud Code of Conduct\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" loading=\"lazy\" decoding=\"async\" srcset=\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png 1200w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=300&amp;h=169 300w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=768&amp;h=432 768w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=1024&amp;h=576 1024w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=264&amp;h=149 264w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=500&amp;h=281 500w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=678&amp;h=381 678w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=150&amp;h=84 150w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=314&amp;h=177 314w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/Salesforce-successfully-declares-Agentforce-adherent-to-the-EU-Cloud-Code-of-Conduct.png?w=343&amp;h=193 343w\" \/>\t\t\t\t\t<\/div>\n\t\n\t<div class=\"content-card__content-container content-card__content-container--style-\">\n\t\t\n\t\t\n\t\t\t\t\t<h3 itemprop=\"headline\" class=\"content-card__title\">\n\t\t\t\t\t\t\t\t\t<a\n\t\t\t\t\t\thref=\"https:\/\/www.salesforce.com\/news\/stories\/agentforce-gdpr-compliance-certification\/\"\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\ttarget=\"_blank\"\n\t\t\t\t\t\t\t\t\t\t\t\tclass=\"content-card__title-link\"\n\t\t\t\t\t>\n\t\t\t\t\t\t\t\tAgentforce Obtains Second-Level Compliance with the EU Cloud Code of Conduct\u00a0\t\t\t\t<\/a>\t\t\t<\/h3>\n\t\t\n\t\t\t\t\n\t\t\n\t\t\t\t\t<div itemprop=\"readTime\" class=\"content-card__read-time\">\n\t\t\t\t1 min read\t\t\t<\/div>\n\t\t\n\t\t\t<\/div>\n\n<\/article>\n\n\n\n<article\t\t\titemscope=\"\"\n\t\titemtype=\"https:\/\/schema.org\/Article\"\n\t\tdata-card-id=\"92621\"\n\t\tclass=\"content-card content-card--large-inline is-entire-area-clickable content-card--is-boxed\"\n\t\t\tdata-clickable-area-link=\"https:\/\/www.salesforce.com\/news\/stories\/how-law-firms-can-lead-the-agentic-ai-era\/\"\n\t>\n\t\t\t<div class=\"content-card__image-container\" itemprop=\"image\" itemscope itemtype=\"https:\/\/schema.org\/ImageObject\">\n\t\t\t<img width=\"1024\" height=\"576\" src=\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=1024\" class=\"content-card__image\" alt=\"How Law Firms Can Lead the Agentic AI Era\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" loading=\"lazy\" decoding=\"async\" srcset=\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png 1200w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=300&amp;h=169 300w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=768&amp;h=432 768w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=1024&amp;h=576 1024w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=264&amp;h=149 264w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=500&amp;h=281 500w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=678&amp;h=381 678w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=150&amp;h=84 150w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=314&amp;h=177 314w, https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2026\/03\/From-Oversight-to-Ownership_-Why-Law-Firms-Must-Lead-the-Agentic-AI-Era.png?w=343&amp;h=193 343w\" \/>\t\t\t\t\t<\/div>\n\t\n\t<div class=\"content-card__content-container content-card__content-container--style-\">\n\t\t\n\t\t\n\t\t\t\t\t<h3 itemprop=\"headline\" class=\"content-card__title\">\n\t\t\t\t\t\t\t\t\t<a\n\t\t\t\t\t\thref=\"https:\/\/www.salesforce.com\/news\/stories\/how-law-firms-can-lead-the-agentic-ai-era\/\"\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\ttarget=\"_blank\"\n\t\t\t\t\t\t\t\t\t\t\t\tclass=\"content-card__title-link\"\n\t\t\t\t\t>\n\t\t\t\t\t\t\t\tHow Law Firms Can Lead the Agentic AI Era \u2014 And What Clients Now Expect\t\t\t\t<\/a>\t\t\t<\/h3>\n\t\t\n\t\t\t\t\n\t\t\n\t\t\t\t\t<div itemprop=\"readTime\" class=\"content-card__read-time\">\n\t\t\t\t8 min read\t\t\t<\/div>\n\t\t\n\t\t\t<\/div>\n\n<\/article>\n\n\t\t<\/div>\n\t\t<noscript>\n<div class=\"pagination-fallback\">\n\n\t\t<a href=\"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/posts\/86736\/page\/2\/?bc=OTH#section-title\">\n\t\t\tOlder Posts\t\t<\/a>\n\t\t<\/div>\n<\/noscript>\n\t<\/div>\n<\/aside>\n\n\n\n<h4 class=\"wp-block-heading\" id=\"h-go-deeper\"><strong>Go deeper:<\/strong><\/h4>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Learn how leaders are <a href=\"https:\/\/www.salesforce.com\/news\/stories\/closing-the-agentic-ai-trust-gap\/\" target=\"_blank\" rel=\"noreferrer noopener\">racing to bridge the AI trust gap<\/a><\/li>\n\n\n\n<li>Discover <a href=\"https:\/\/www.salesforce.com\/news\/stories\/ai-trust-patterns\/\" target=\"_blank\" rel=\"noreferrer noopener\">how Salesforce builds trust and guardrails into its AI products<\/a><\/li>\n\n\n\n<li>Consider this <a href=\"https:\/\/www.salesforce.com\/news\/stories\/agentic-maturity-model\/\" target=\"_blank\" rel=\"noreferrer noopener\">4-stage maturity model for agentic success<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Key Takeaways AI large language models (LLMs) hallucinate when they generate false but&nbsp; \u200cplausible-sounding responses based on flawed or incomplete data. It\u2019s unintentional. More fiction than fraud. But when AI knows the truth and chooses not to tell you, that\u2019s different. It\u2019s engaging in deception. Not because the system is sinister, like something out of [&hellip;]<\/p>\n","protected":false},"author":490,"featured_media":86746,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"sf_subhead":"","sf_i18n_disclaimer":false,"_jetpack_memberships_contains_paid_content":false,"alternateThumbnailId":0,"sf_product_cta_id":0,"footnotes":""},"categories":[1],"tags":[],"sf_content_type":[1730],"sf_theme":[1916],"sf_topic":[20528,1737,2290,1723],"sf_product":[19930,2717,1748],"sf_industry":[],"sf_role":[],"sf_multimedia_asset":[],"sf_location":[1798],"sf_collection":[21199],"sf_visibility":[],"coauthors":[18423],"class_list":["post-86736","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized","sf_content_type-feature-stories","sf_theme-artificial-intelligence","sf_topic-agents","sf_topic-artificial-intelligence","sf_topic-technology-ethics","sf_topic-trust","sf_product-agentforce","sf_product-data-360","sf_product-salesforce-platform","sf_location-global","sf_collection-fy26-q2"],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v27.2 (Yoast SEO v27.2) - https:\/\/yoast.com\/product\/yoast-seo-premium-wordpress\/ -->\n<title>Does AI Lie? Salesforce Weighs in. - Salesforce<\/title>\n<meta name=\"description\" content=\"Key Takeaways An AI model can engage in &quot;deceptive alignment,&quot; secretly pursuing its own goals instead of developers&#039; stated objectives. This can manifest\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"When AI &#039;Lies,&#039; Trust and Guardrails Are Even More Critical\" \/>\n<meta property=\"og:description\" content=\"Key Takeaways An AI model can engage in &quot;deceptive alignment,&quot; secretly pursuing its own goals instead of developers&#039; stated objectives. This can manifest\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\" \/>\n<meta property=\"og:site_name\" content=\"Salesforce\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/salesforce\/\" \/>\n<meta property=\"article:published_time\" content=\"2025-05-02T15:00:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-09-23T15:33:15+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.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=\"Dave Berkowitz\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@salesforcenews\" \/>\n<meta name=\"twitter:site\" content=\"@salesforcenews\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Dave Berkowitz\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\"},\"author\":[{\"@id\":\"https:\/\/www.salesforce.com\/news\/#\/schema\/person\/image\/ddfc697f11cd55f34758e00eb5b0cfd4\"}],\"headline\":\"When AI &#8216;Lies,&#8217; Trust and Guardrails Are Even More Critical\",\"datePublished\":\"2025-05-02T15:00:00+00:00\",\"dateModified\":\"2025-09-23T15:33:15+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\"},\"wordCount\":1394,\"publisher\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/#organization\"},\"image\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png\",\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\",\"url\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\",\"name\":\"Does AI Lie? Salesforce Weighs in. - Salesforce\",\"isPartOf\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png\",\"datePublished\":\"2025-05-02T15:00:00+00:00\",\"dateModified\":\"2025-09-23T15:33:15+00:00\",\"description\":\"Key Takeaways An AI model can engage in \\\"deceptive alignment,\\\" secretly pursuing its own goals instead of developers' stated objectives. This can manifest\",\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage\",\"url\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png\",\"contentUrl\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png\",\"width\":1200,\"height\":675,\"caption\":\"Does AI lie?\"},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.salesforce.com\/news\/#website\",\"url\":\"https:\/\/www.salesforce.com\/news\/\",\"name\":\"Salesforce\",\"description\":\"Get the latest Salesforce press releases, announcements, stories, and media contacts. See today\u2019s CRM news.\",\"publisher\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.salesforce.com\/news\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.salesforce.com\/news\/#organization\",\"name\":\"Salesforce\",\"url\":\"https:\/\/www.salesforce.com\/news\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.salesforce.com\/news\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2020\/08\/default.jpg\",\"contentUrl\":\"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2020\/08\/default.jpg\",\"width\":1200,\"height\":630,\"caption\":\"Salesforce\"},\"image\":{\"@id\":\"https:\/\/www.salesforce.com\/news\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/salesforce\/\",\"https:\/\/x.com\/salesforcenews\",\"https:\/\/www.linkedin.com\/company\/salesforce\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.salesforce.com\/news\/#\/schema\/person\/image\/ddfc697f11cd55f34758e00eb5b0cfd4\",\"name\":\"Dave Berkowitz\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g23494c9101089ad44ae88ce9d2f56aac\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g\",\"caption\":\"Dave Berkowitz\"},\"url\":\"https:\/\/www.salesforce.com\/news\/stories\/author\/dave-berkowitz\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Does AI Lie? Salesforce Weighs in. - Salesforce","description":"Key Takeaways An AI model can engage in \"deceptive alignment,\" secretly pursuing its own goals instead of developers' stated objectives. This can manifest","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:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/","og_locale":"en_US","og_type":"article","og_title":"When AI 'Lies,' Trust and Guardrails Are Even More Critical","og_description":"Key Takeaways An AI model can engage in \"deceptive alignment,\" secretly pursuing its own goals instead of developers' stated objectives. This can manifest","og_url":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/","og_site_name":"Salesforce","article_publisher":"https:\/\/www.facebook.com\/salesforce\/","article_published_time":"2025-05-02T15:00:00+00:00","article_modified_time":"2025-09-23T15:33:15+00:00","og_image":[{"width":1200,"height":675,"url":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","type":"image\/png"}],"author":"Dave Berkowitz","twitter_card":"summary_large_image","twitter_creator":"@salesforcenews","twitter_site":"@salesforcenews","twitter_misc":{"Written by":"Dave Berkowitz","Est. reading time":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#article","isPartOf":{"@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/"},"author":[{"@id":"https:\/\/www.salesforce.com\/news\/#\/schema\/person\/image\/ddfc697f11cd55f34758e00eb5b0cfd4"}],"headline":"When AI &#8216;Lies,&#8217; Trust and Guardrails Are Even More Critical","datePublished":"2025-05-02T15:00:00+00:00","dateModified":"2025-09-23T15:33:15+00:00","mainEntityOfPage":{"@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/"},"wordCount":1394,"publisher":{"@id":"https:\/\/www.salesforce.com\/news\/#organization"},"image":{"@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage"},"thumbnailUrl":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/","url":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/","name":"Does AI Lie? Salesforce Weighs in. - Salesforce","isPartOf":{"@id":"https:\/\/www.salesforce.com\/news\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage"},"image":{"@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage"},"thumbnailUrl":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","datePublished":"2025-05-02T15:00:00+00:00","dateModified":"2025-09-23T15:33:15+00:00","description":"Key Takeaways An AI model can engage in \"deceptive alignment,\" secretly pursuing its own goals instead of developers' stated objectives. This can manifest","inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.salesforce.com\/news\/stories\/does-ai-lie\/#primaryimage","url":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","contentUrl":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","width":1200,"height":675,"caption":"Does AI lie?"},{"@type":"WebSite","@id":"https:\/\/www.salesforce.com\/news\/#website","url":"https:\/\/www.salesforce.com\/news\/","name":"Salesforce","description":"Get the latest Salesforce press releases, announcements, stories, and media contacts. See today\u2019s CRM news.","publisher":{"@id":"https:\/\/www.salesforce.com\/news\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.salesforce.com\/news\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.salesforce.com\/news\/#organization","name":"Salesforce","url":"https:\/\/www.salesforce.com\/news\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.salesforce.com\/news\/#\/schema\/logo\/image\/","url":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2020\/08\/default.jpg","contentUrl":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2020\/08\/default.jpg","width":1200,"height":630,"caption":"Salesforce"},"image":{"@id":"https:\/\/www.salesforce.com\/news\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/salesforce\/","https:\/\/x.com\/salesforcenews","https:\/\/www.linkedin.com\/company\/salesforce\/"]},{"@type":"Person","@id":"https:\/\/www.salesforce.com\/news\/#\/schema\/person\/image\/ddfc697f11cd55f34758e00eb5b0cfd4","name":"Dave Berkowitz","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g23494c9101089ad44ae88ce9d2f56aac","url":"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/?s=96&d=mm&r=g","caption":"Dave Berkowitz"},"url":"https:\/\/www.salesforce.com\/news\/stories\/author\/dave-berkowitz\/"}]}},"jetpack_featured_media_url":"https:\/\/www.salesforce.com\/news\/wp-content\/uploads\/sites\/3\/2025\/05\/When-AI-\u2018Lies-Trust-and-Guardrails-are-Even-More-Critical.png","jetpack_sharing_enabled":true,"distributor_meta":false,"distributor_terms":false,"distributor_media":false,"distributor_original_site_name":"Salesforce","distributor_original_site_url":"https:\/\/www.salesforce.com\/news","push-errors":false,"_links":{"self":[{"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/posts\/86736","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/users\/490"}],"replies":[{"embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/comments?post=86736"}],"version-history":[{"count":6,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/posts\/86736\/revisions"}],"predecessor-version":[{"id":89719,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/posts\/86736\/revisions\/89719"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/media\/86746"}],"wp:attachment":[{"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/media?parent=86736"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/categories?post=86736"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/tags?post=86736"},{"taxonomy":"sf_content_type","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_content_type?post=86736"},{"taxonomy":"sf_theme","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_theme?post=86736"},{"taxonomy":"sf_topic","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_topic?post=86736"},{"taxonomy":"sf_product","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_product?post=86736"},{"taxonomy":"sf_industry","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_industry?post=86736"},{"taxonomy":"sf_role","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_role?post=86736"},{"taxonomy":"sf_multimedia_asset","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_multimedia_asset?post=86736"},{"taxonomy":"sf_location","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_location?post=86736"},{"taxonomy":"sf_collection","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_collection?post=86736"},{"taxonomy":"sf_visibility","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/sf_visibility?post=86736"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.salesforce.com\/news\/wp-json\/wp\/v2\/coauthors?post=86736"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}