The Jurisprudential Evolution of Agentic Commerce
A Comparative Analysis of Legal Frameworks and Litigation Trends in the US, UK, and EU
In November 2025, Amazon sued Perplexity AI in the Northern District of California over Comet, Perplexity’s AI browser agent. The case raises a deceptively simple question: can a user authorise an AI agent to access a password-protected marketplace on their behalf where the platform owner has expressly prohibited automated access? On 9 March 2026, Judge Maxine Chesney answered that question at the preliminary injunction stage in Amazon’s favour, holding that Amazon had shown strong evidence of unauthorised access. The Ninth Circuit’s subsequent administrative stay means the point is not yet settled. But the litigation has already framed the central legal issue for agentic commerce: the difference between user authority and platform authorisation.5,6,17
The Amazon v. Perplexity litigation matters because it is beginning to articulate the boundary between user authority and platform control. Globally, the commercial architecture is shifting from traditional e-commerce to agentic commerce. This transition is marked by the delegation of consumer decisions to autonomous AI systems.1,2 This shift moves beyond passive assistance to a state where AI agents sense their environment, plan multi-step workflows and execute transactions with little to no human supervision.1,2 McKinsey research suggests that by 2030, agentic commerce could orchestrate between three trillion and five trillion dollars in global revenue, with the US retail sector alone accounting for up to one trillion.3,4 However, this rapid integration of autonomous agents into commerce has outpaced existing legal structures, creating a “regulatory grey zone” across the United States, the United Kingdom and the European Union.3 The Amazon litigation now provides one of the first major attempts to map the boundaries of that zone.
The Conceptual Architecture of Agentic Commerce
To understand the legal challenges posed by agentic commerce, one must first distinguish these systems from earlier forms of automation. While earlier systems followed predefined, rule-based processes, agentic AI uses Large Language Models (LLMs) and reasoning frameworks to pursue high-level objectives over time.2,7 These systems possess four core capabilities: autonomy from continuous supervision; goal-orientation; multi-step reasoning across disparate data sources; and the ability to act across multiple services or platforms.2,7
In a commercial context, agentic systems operate at three levels of increasing sophistication. At the first level, agents interact directly with merchant platforms, mimicking human browsing to find and purchase goods. This pattern is termed “agent-to-site” transactions.1 At the second level, systems transact with other agents, negotiating terms without human intermediaries.1,8 At the third level, intermediary systems coordinate complex interactions across a global network of platforms, creating what may be called “multi-agent ecosystems.”1
This shift moves the consumer from “using tools” to “delegating outcomes.”2,9 Legally, this creates a “third actor problem.” The traditional bilateral relationship between consumer and merchant is complicated by an autonomous intermediary that may lack legal personality but can bind its principal to contracts, often without the principal knowing the specific terms.3,10
US Litigation and the Precedent of Platform Authorisation
In the United States, the legal position on agentic commerce is being shaped primarily through the Computer Fraud and Abuse Act of 1986 (CFAA), a statute originally drafted to combat hacking that is now being tested against autonomous shopping agents.11,12,13
The Test Case: Amazon.com Services LLC v. Perplexity AI, Inc.
In November 2025, Amazon filed suit against Perplexity AI in the Northern District of California, alleging that the developer’s “Comet” browser agent was configured to access password-protected Amazon customer accounts to browse products and complete purchases.5,6,16 According to Amazon’s complaint, the agent deliberately concealed its automated nature by spoofing its User-Agent string to appear as a standard Google Chrome session, thereby evading Amazon’s bot detection systems.5,14,15,16 Amazon contended that it issued at least five warnings and deployed technical barriers, which Perplexity bypassed within 24 hours.5,16
Judge Maxine Chesney’s preliminary injunction rests on several key legal findings. First, the court ruled that user permission is not legally equivalent to platform authorisation. While a user may consent to an AI agent using their credentials, the platform owner retains the right to refuse that agent access to its infrastructure.6,18 Second, the court found that by bypassing technical blocks and disguising its identity, Perplexity likely violated the CFAA’s prohibition on accessing a protected computer “without authorization.”6,19,20 Third, the injunction rested partly on evidence that uncontrolled AI agents interfere with Amazon’s advertising systems, which require accurate detection of automated traffic to maintain billing integrity for advertisers.14
The Ninth Circuit Court of Appeals granted an emergency administrative stay of the injunction on March 16, 2026, allowing Perplexity to continue operations while the merits of the appeal are considered.24,25,26 Perplexity’s defence argues that the CFAA does not prohibit accessing public websites and that the “access” in question is performed by the human consumer, not the company.25
Evolution of the CFAA and Parallel Scraper Litigation
Amazon v. Perplexity represents the next step in a line of scraper cases. In hiQ Labs, Inc. v. LinkedIn Corp., the Ninth Circuit established that scraping publicly available data (where no login is required) does not constitute unauthorised access under the CFAA.12,20,27 The Amazon case shifts the focus to “gated” environments. The court cited Facebook, Inc. v. Power Ventures, Inc., which established that once a platform owner explicitly withdraws permission via a cease-and-desist letter, any further access constitutes a CFAA violation even if account holders provide their credentials.6,11,20 The distinction is not simply scraping versus non-scraping; it is public web access versus access to account-protected or technically restricted environments after permission has been withdrawn.
A further parallel is Ryanair v. Booking.com. In July 2024, a Delaware jury found that Booking.com violated the CFAA with an “intent to defraud” by screen scraping the airline’s website to resell tickets with unauthorised fees.28,29 Ryanair’s win is consistent with a line of cases in which courts have been more receptive to platform-control arguments where access is technically restricted, commercial or contrary to express withdrawal of permission.28,29
Privacy and Consumer Protection Claims in the US
Litigation in the United States is also extending into privacy and background-check statutes. In Ambriz v. Google LLC, plaintiffs allege that Google’s AI-powered Cloud Contact Center functions as a “third-party listener” that analyses call data without proper consent.11,12 A separate class action filed in January 2026 targets Eightfold AI, alleging that its AI-powered hiring tools violate the Fair Credit Reporting Act (FCRA) and California’s Investigative Consumer Reporting Agencies Act (ICRAA).30 The complaint argues that by assembling individualised reports from diverse internet sources to assess job suitability, the AI service becomes a “consumer reporting agency” subject to strict disclosure and accuracy requirements.30
US courts have also begun sanctioning lawyers for submitting AI-generated filings that include fabricated case citations, as seen in Mata v. Avianca, Inc..31,32,33 These rulings confirm that AI is a tool and that human professionals bear ultimate responsibility for the accuracy of information submitted under their name.31,33
The UK Approach: Sectoral Regulation and Business Responsibility
The United Kingdom has avoided the EU’s path of horizontal statutory AI regulation, opting instead for a “pro-innovation,” sectoral approach.34,35,36,37 Oversight is distributed among existing regulators: the Competition and Markets Authority (CMA), the Information Commissioner’s Office (ICO), and the Financial Conduct Authority (FCA).9,36,38
CMA Guidance: Direct Accountability for AI Agents
On March 9, 2026, the CMA published detailed guidance on complying with consumer law when using AI agents.9 The regulator’s core position is that businesses cannot outsource their legal responsibilities to autonomous code; they are as accountable for the actions of their AI agents as they are for those of their human employees.9,39
The CMA’s guidance can be reduced to three practical compliance duties. First, businesses must embed consumer protection principles into agent design, ensuring that agents respect statutory rights, provide accurate pricing, and avoid misleading omissions.9,39,40 Second, businesses need appropriate human oversight and monitoring to catch hallucinations, inaccurate outputs and other failures that could mislead consumers or cause financial loss.9,39,40 Third, if an agent is found to be non-compliant or making errors, the business must act immediately to refine prompts or workflows.9,40 Failure to correct an identified breach promptly will result in accountability.
Under the Digital Markets, Competition and Consumer Act 2024 (DMCC Act), the CMA has the power to impose administrative fines of up to 10 per cent of a company’s global turnover for consumer law breaches, including misleading sales practices delivered by AI systems.9,41
English Common Law and AI Liability
The UK Jurisdiction Taskforce (UKJT) launched a consultation in January 2026 on a draft Legal Statement on liability for AI harms under English private law.42,43,44 The statement holds that under English law, AI does not have legal personality and therefore cannot be held responsible for physical or economic harm.44,45,46 Liability must instead be attributed to legal persons through existing principles of negligence.44,45,46
The UKJT analysis examines vicarious liability, asking whether an employer may be vicariously liable for an employee’s negligent use of AI, though generally not for the AI system itself.45,46 It also considers professional standard of care, asking whether a professional may be liable for breach of duty by failing to perform due diligence on an AI system before using it for client work, and conversely, whether failure to use AI where a reasonable professional would do so might also constitute a breach.45,46 The analysis further addresses factual causation. The “black box” nature of AI, where the internal reasoning behind a specific output is opaque, may require courts to approach causation differently through evidential assumptions or expert testimony.46,47 Because the UKJT statement remains a draft consultation document, it should be treated as influential guidance on likely common-law analysis, not as a binding statement of law.
Hallucinations in UK Courts and Tribunals
By early 2026, the UK legal system had recorded at least 31 reported instances of AI hallucinations in litigation, including fictitious case citations and misrepresented legal norms.48 In Taiwo v. Homelets of Bath Ltd (2025), the High Court issued a warning after a litigant submitted a skeleton argument referencing cases that did not exist.49,50 The managing partner in Choksi v. IPS Law LLP (2025) submitted a witness statement found to contain fabricated authorities, leading to costs orders and potential referrals to regulators.50,51 In Hassan v. ABC International Bank, a tribunal found 46 inaccurate citations (9 fictitious and 37 misrepresentations) and concluded the conduct was reckless and unreasonable, justifying a costs order.50
The judiciary has updated its AI Guidance for Judicial Office Holders twice in 2025 to address these risks, emphasising that confidentiality must be maintained and that AI use must protect the integrity of the administration of justice.48,52,53
The European Union: The Horizontal Statutory Model
In Brussels, the EU has enacted the world’s first horizontal law regulating Artificial Intelligence: the EU AI Act.54,55,56,57 While the Act does not specifically name “agentic AI,” its definition of AI systems focuses on machine-based systems that operate with varying levels of autonomy and adaptiveness. This definition clearly encompasses autonomous agents.58,59,60,61
The AI Act and High-Risk Agentic Systems
Under the EU AI Act’s tiered, risk-based approach, some agentic systems used in commerce may be high-risk, but only where they fall within the Act’s product-safety route or one of the Annex III use cases, such as creditworthiness, employment, access to essential services or other decisions affecting fundamental rights.54,61,64 Ordinary retail discovery or checkout agents should not be described as high-risk merely because they are autonomous. One unresolved question is whether “human control” must be exercised in real-time at the moment of an autonomous purchase, or whether setting parameters in advance constitutes sufficient oversight.3,54
Liability and the Revised Product Liability Directive
The Commission officially withdrew the proposed AI Liability Directive in 2025. The revised Product Liability Directive (PLD) now does part of the practical work that the AI liability package was expected to perform: it brings software, including AI systems, within the concept of a product and updates disclosure and evidential rules for technically complex claims.63,67,68,69,70 It should not, however, be described as a full replacement for the withdrawn AI Liability Directive.
Under the revised PLD, strict liability applies: manufacturers are liable for damages caused by product defects.69,70 In cases of extreme technical complexity, courts may mandate disclosure of evidence and create a rebuttable presumption that the AI system was defective.67 Importers from third countries are held to the same standards where their AI outputs are used within the Union.67
Some commentators and payments-sector observers have suggested a future category of “supervised digital agents,” sitting between human users and autonomous machine actors.3 That idea is not yet a settled EU legislative category and should be presented as a possible future governance model rather than current law. The concept would formalise delegation credentials and tamper-evident behavioural logs for autonomous action systems.3,67
GDPR and CJEU Clarifications on Automated Decision-Making
Under the General Data Protection Regulation (GDPR), AI agents processing personal data remain subject to Article 22, which provides individuals the right not to be subject to a decision based “solely on automated processing” that produces legal consequences or substantially affects their interests.37,66,71,72
Recent CJEU case law has sharpened these rights. In the Schufa Case, the Court ruled that an automated creditworthiness score constitutes an automated “decision” if it plays a decisive role in a lender’s choice, even if a human nominally makes the final determination.67,71 In the D&B Case, the Court held that meaningful information about the logic involved must be provided in an intelligible form, while trade-secret concerns may require procedural balancing rather than simple refusal.71,73 Consumers also retain the right to challenge AI-driven outcomes and request human review whenever a decision may affect them.74
These rulings indicate that as AI agents take on more commerce tasks (insurance risk assessment, dynamic pricing), they will face more demanding transparency and explainability requirements.58,61,73
Systemic Effects and Strategic Implications of Agentic Commerce
Agentic commerce marks a turning point for the web, as it shifts from a human-readable medium to one that is machine-actionable.4,75,76
The End of Permissionless Scraping and the Rise of AEO
The Amazon v. Perplexity ruling, if upheld on appeal, signals the beginning of the end for permissionless AI access to gated commercial platforms.75 For years, parts of web scraping operated in a legal grey zone, but the court’s endorsement of platform control suggests that the practice of quietly extracting data and seeking forgiveness afterwards is drawing to a close.22,77
For merchants, this requires a shift from Search Engine Optimization (SEO) to “Answer Engine Optimization” (AEO).3,78 Brands must now make their product data, pricing, and return policies machine-readable so that AI agents can find, evaluate, and transact with them.3,18,76 Cloudflare’s “Markdown for Agents” is an early technical response to this trend, allowing enabled sites to serve Markdown versions of HTML pages to AI agents through content negotiation.76
The Race for Standardised Transactional Protocols
A competition is under way to define the communication standards for agentic commerce. Two primary models are emerging. Google’s Universal Commerce Protocol (UCP), launched in early 2026, establishes a common language between AI agents, merchant systems, and payment providers.18,79,80,81 Google says UCP was co-developed with Shopify, Etsy, Wayfair, Target and Walmart and endorsed by more than 20 ecosystem participants, including Mastercard, Visa, Stripe, Adyen and The Home Depot.80,81 It allows agents to identify themselves and transact through infrastructure that the retailer controls.18,81
The second model is OpenAI and Stripe’s Agentic Commerce Protocol (ACP), which embeds the entire shopping journey within conversational interfaces like ChatGPT, with Instacart already offering “Instant Checkout” features via ACP.79,80
Both protocols aim to reduce the friction that leads to abandoned carts while giving merchants a way to monetise bot traffic that might otherwise bypass traditional advertising surfaces.79,82,83,84
Consumer Agency vs. Manipulative “Dark Patterns”
While agentic commerce offers genuine time savings and reduced decision fatigue for consumers, it also introduces new risks.2 Consumer advocacy groups like Bureau Européen des Unions de Consommateurs (European Consumer Organisation) have warned that AI agents could steer users toward outcomes that benefit the developers or the retailers paying the highest commissions, rather than the consumer’s actual interest.85,86,87
The CMA and European Commission are particularly concerned about “dark patterns” (interface designs that mislead consumers into decisions they would not otherwise make, such as false countdown timers or manufactured scarcity messaging).39,88,89 In the agentic era, these patterns could become invisible to the human user, occurring as “agent-to-agent” interactions where one bot deceives another.88
Algorithmic collusion also merits attention. Competition authorities in the UK and EU have made algorithmic pricing a priority.90,91 The concern is that if multiple competitors use the same pricing algorithm or share commercially sensitive data through a common third-party provider, it could produce coordinated price increases without any explicit human communication. This phenomenon is termed “agentic collusion.”9,90,91,92 In February 2026, the CMA launched an investigation into hotel chains suspected of sharing information through a common data services provider.90 Polish competition authorities confirmed multiple ongoing investigations into collusion via algorithmic tools in the banking and pharmaceutical sectors in late 2025.90 In the United States, Agri Stats litigation illustrates the same hub-and-spoke concern in a data-services setting, even though it is not itself an AI-agent case. In January 2026, Agri Stats agreed to resolve a meat-workers wage-suppression class action, and in May 2026 the DOJ and several states announced a proposed settlement of their meat-pricing case.90,93
Conclusions and Strategic Implications
The transition to agentic commerce exhibits a widening gap between what the technology can do and what the law has settled on.3 The Amazon v. Perplexity litigation has, at least at the preliminary injunction stage, framed the central platform-access issue for agentic commerce: user-delegated authority is not necessarily the same thing as platform authorisation.6,18 As agentic systems enter commercial deployment, the question before courts and regulators is shifting from “is it possible?” to “who is responsible when it fails?”3,95,96
For stakeholders in the digital economy, four priorities warrant attention. First, AI developers should move toward “permissioned AI” architectures that secure both user credentials and explicit platform authorisation via official APIs. This “Dual Authorisation” model stands in contrast to the user-permission-only model that Perplexity attempted.20,75,81
Second, organisations must implement regularly updated compliance records and detailed audit logs as required by the EU AI Act and recommended by the CMA.9,35,40,66,97 This includes real-time monitoring of agent behaviour to detect hallucinations, bias and potential collusive patterns before they generate serious liability.9,39,97 Such “Governance by Design” embeds accountability into the systems themselves rather than relying on post-facto enforcement.
Third, businesses should be transparent when using AI agents, particularly where a consumer’s knowledge that they are dealing with an AI (rather than a human) would affect their decision to proceed.9,39,40 “Transparent Consumer Disclosure” is not merely a compliance obligation; it also respects the autonomy of individuals making delegated purchasing decisions through agentic systems.
Fourth, companies should review and update technology contracts to address agentic transactions, defining the scope of authority granted to AI and setting out mechanisms for dispute resolution and indemnification.1,98 Standard disclaimers written for passive software may no longer be legally adequate for autonomous systems that execute high-value contracts.98 This “Contractual Risk Allocation” approach forces organisations to grapple explicitly with the liability surface created when machines bind their principals to commercial obligations.
As the global market for agentic commerce grows toward five trillion dollars by 2030, the legal frameworks of the United States, United Kingdom and European Union will continue to be tested.3,4 The cases decided over the next two to three years will determine whether agentic commerce develops on a permissioned, regulated basis or whether the law trails the technology by a decade, as it has before.
References
1 Digital commerce redefined: The growing impact of agentic commerce --- Linklaters / TechInsights. https://techinsights.linklaters.com/post/102lwgk/digital-commerce-redefined-the-growing-impact-of-agentic-commerce
2 Agentic AI and consumers --- GOV.UK. https://www.gov.uk/government/publications/agentic-ai-and-consumers/agentic-ai-and-consumers
3 Agentic Commerce: When AI Buys on Your Behalf --- European Business Magazine. https://europeanbusinessmagazine.com/business/agentic-commerce-when-ai-buys-on-your-behalf-who-pays-whos-liable/
4 The agentic commerce opportunity --- McKinsey. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-agentic-commerce-opportunity-how-ai-agents-are-ushering-in-a-new-era-for-consumers-and-merchants
5 Judge blocks Perplexity’s AI bot from shopping on Amazon --- GeekWire. https://www.geekwire.com/2026/judge-blocks-perplexitys-ai-bot-from-shopping-on-amazon-in-early-test-of-agentic-commerce/
6 Court Finds AI Agent May Violate Federal Law by Accessing Amazon Accounts Without Authorization --- Cooley. https://www.cooley.com/news/insight/2026/2026-03-17-court-finds-ai-agent-may-violate-state-federal-law-by-accessing-amazon-accounts-without-authorization
7 Agentic AI in Financial Services: Regulatory and Legal Considerations --- Hogan Lovells. https://www.hoganlovells.com/en/publications/agentic-ai-in-financial-services-regulatory-and-legal-considerations
8 Amazon vs. Perplexity: A Lawsuit Over AI Agents that Shop --- Marketing AI Institute. https://www.marketingaiinstitute.com/blog/amazon-sues-perplexity
9 Free agent? New UK guidance on agentic AI for businesses --- Ashurst. https://www.ashurst.com/en/insights/free-agent-not-quite-new-uk-guidance-on-agentic-ai-for-businesses/
10 Agentic commerce: can the law of contract adapt? --- Gilbert + Tobin. https://www.gtlaw.com.au/insights/agentic-commerce-can-the-law-of-contract-adapt
11 How 2 Tech Statutes Are Being Applied To Agentic AI --- Weil, Gotshal & Manges LLP. https://www.weil.com/-/media/files/pdfs/2026/february/law360\--how-2-tech-statutes-are-being-applied-to-agentic-ai.pdf
12 AI Agents Are Raising New Questions of Fraud and Privacy Liability --- PYMNTS.com. https://www.pymnts.com/cpi-posts/ai-agents-are-raising-new-questions-of-fraud-and-privacy-liability/
13 Computer Fraud and Abuse Act (CFAA) --- NACDL. https://www.nacdl.org/Landing/ComputerFraudandAbuseAct
14 Amazon wins order blocking access by Perplexity’s AI shopping agent --- Reuters. https://www.reuters.com/legal/litigation/amazon-wins-order-blocking-access-perplexitys-ai-shopping-agent-2026-03-10/
15 Federal judge blocks Perplexity’s AI browser from making Amazon purchases --- CyberScoop. https://cyberscoop.com/amazon-perplexity-comet-browser-injunction/
16 Amazon vs. Perplexity AI: Court Blocks Comet Shopping Agent (2026) --- ALM Corp. https://almcorp.com/blog/amazon-vs-perplexity-ai-comet-shopping-agent-court-ruling/
17 Court Blocks Perplexity from Accessing Amazon in Agentic AI Lawsuit --- The Fashion Law. https://www.thefashionlaw.com/court-blocks-perplexity-from-accessing-amazon-systems-in-agentic-ai-lawsuit/
18 The Authorization Gap: What Amazon’s Perplexity Injunction Reveals --- Winning with Walmart. https://winningwithwalmart.com/the-authorization-gap-what-amazons-perplexity-injunction-reveals-about-the-commerce-layer-walmart-is-building-differently/
19 Amazon.com Services LLC v. Perplexity AI Inc., No. 3:25-cv-09514, Preliminary Injunction Order (N.D. Cal. Mar. 9, 2026).
20 hiQ Labs, Inc. v. LinkedIn Corp., 31 F.4th 1180 (9th Cir. 2022); Facebook, Inc. v. Power Ventures, Inc., 844 F.3d 1058 (9th Cir. 2016).
21 Computer Fraud and Abuse Act Update: Supreme Court decides Van Buren v. U.S. https://hh-law.com/blogs/blog-intellectual-property-litigation-protection-and-prosecution-dtsa-ai-artificial-intelligence-lawyers/cfaa-van-buren-v-united-states/
22 Van Buren v. United States, 593 U.S. 374 (2021).
23 Amazon.com Services LLC v. Perplexity AI Inc., Complaint, No. 3:25-cv-09514 (N.D. Cal. filed Nov. 2025).
24 Appeals court temporarily pauses order blocking Perplexity’s AI shopping agent on Amazon --- CyberScoop. https://cyberscoop.com/perplexity-comet-ai-shopping-agent-amazon-lawsuit-ninth-circuit-stay/
25 Court Temporarily Lifts Order Banning Perplexity From Amazon --- MediaPost. https://www.mediapost.com/publications/article/413555/court-temporarily-lifts-order-banning-perplexity-f.html
26 Court temporarily allows Perplexity AI shopping ‘agents’ on Amazon --- Westlaw Today. https://today.westlaw.com/Document/Ia1b3afe0222811f1bf7dc89c10580585/View/FullText.html
27 The Legal Landscape of Web Scraping --- Quinn Emanuel. https://www.quinnemanuel.com/the-firm/publications/the-legal-landscape-of-web-scraping/
28 Ryanair Wins Legal Case Against Booking.com Over Screen Scraping --- Simple Flying. https://simpleflying.com/ryanair-wins-case-booking-screen-scraping-reselling-tickets/
29 Ryanair wins Booking.com ‘screen scraper’ case --- Aviation Business News. https://www.aviationbusinessnews.com/industry-news/ryanair-wins-booking-com-screen-scraper-case/
30 Groundbreaking Lawsuit Tests Whether AI Hiring Tools Trigger FCRA Compliance --- Ogletree. https://ogletree.com/insights-resources/blog-posts/groundbreaking-lawsuit-tests-whether-ai-hiring-tools-trigger-fcra-compliance/
31 Lawyer Fined for Using AI-Generated Legal Documents with Fake Citations --- Spellbook. https://www.spellbook.legal/learn/lawyer-fined-using-ai-legal-fake-citations
32 Mata v. Avianca, Inc., No. 22-cv-1461 (S.D.N.Y. 2023), sanctions order.
33 Federal Court Rules Client’s AI-Generated Documents Not Privileged --- JD Supra. https://www.jdsupra.com/legalnews/federal-court-rules-client-s-ai-6386760/
34 The Future of Legal AI in the UK: Trends to Watch in 2025 --- Qanooni. https://qanooni.ai/blog/the-future-of-legal-ai-in-the-uk-trends-to-watch-in-2025/
35 AI regulation: a comparative overview of the UK, EU and US --- Stevens & Bolton. https://www.stevens-bolton.com/insights/102kd49/ai-regulation-a-comparative-overview-of-the-uk-eu-and-us/
36 AI regulation in the UK: The role of the regulators --- Bird & Bird. https://www.twobirds.com/en/insights/2026/uk/ai-regulation-in-the-uk-the-role-of-the-regulators
37 Agentic AI in the workplace --- Prettys Solicitors LLP. https://prettys.co.uk/articles/agentic-ai-in-the-workplace/
38 AI and the FCA: our approach. https://www.fca.org.uk/firms/innovation/ai-approach
39 Complying with consumer law when using AI agents --- GOV.UK. https://www.gov.uk/government/publications/complying-with-consumer-law-when-using-ai-agents/complying-with-consumer-law-when-using-ai-agents
40 Agentic AI and consumer law: the CMA’s guidance for businesses --- Lewis Silkin LLP. https://www.lewissilkin.com/insights/2026/03/13/agentic-ai-and-consumer-law-the-cmas-guidance-for-businesses-102mmud
41 CMA launches first DMCCA enforcement cases --- RPC Legal. https://www.rpclegal.com/snapshots/consumer/winter-2025/cma-launches-first-dmcca-enforcement-cases-and-finalises-price-transparency-guidance/
42 UK Jurisdiction Taskforce launches consultation on liability for AI harms --- Henderson Chambers. https://www.hendersonchambers.co.uk/2026/01/22/uk-jurisdiction-taskforce-launches-consultation-on-liability-for-ai-harms/
43 UK Jurisdiction Taskforce launches consultation on liability for AI harms --- Practical Law. https://uk.practicallaw.thomsonreuters.com/w-049-2237
44 UKJT publishes consultation on its Legal Statement on Liability for AI Harms --- Bird & Bird. https://www.twobirds.com/en/insights/2026/uk/uk-jurisdiction-taskforce-publishes-consultation-on-its-legal-statement-on-liability-for-ai-harms
45 UKJT consultation: Liability for AI harms under English private law --- CMS. https://cms.law/en/gbr/legal-updates/ukjt-consultation-liability-for-ai-harms-under-english-private-law
46 UK Jurisdiction Taskforce consults on draft legal statement on liability for AI harms --- HSF Kramer. https://www.hsfkramer.com/notes/litigation/2026-01/uk-jurisdiction-taskforce-consults-on-draft-legal-statement-on-liability-for-ai-harms
47 What are the litigation trends in 2026? --- Taylor Wessing. https://www.taylorwessing.com/en/insights-and-events/insights/2026/03/what-are-the-litigation-trends-in-2026
48 Fake cases... Will hallucinations stop? --- Counsel Magazine. https://www.counselmagazine.co.uk/articles/fake-cases-will-hallucinations-stop-
49 Taiwo v. Homelets of Bath Ltd \[2025\] EWHC 3173 --- Leathes Prior. https://www.leathesprior.co.uk/news/taiwo-v-homelets-of-bath-ltd-ors-2025-ewhc-3173-kb-03-december-2025
50 AI Hallucination Cases Database --- Damien Charlotin. https://www.damiencharlotin.com/hallucinations/
51 The increasing legal liability of AI hallucinations --- VinciWorks. https://vinciworks.com/blog/the-increasing-legal-liability-of-ai-hallucinations-why-uk-law-firms-face-rising-regulatory-and-litigation-risk/
52 Artificial Intelligence (AI) --- Judicial Guidance (October 2025). https://www.judiciary.uk/guidance-and-resources/artificial-intelligence-ai-judicial-guidance-october-2025/
53 Key developments in AI disputes in 2025 --- Taylor Wessing. https://www.taylorwessing.com/en/insights-and-events/insights/2026/01/key-developments-in-ai-disputes-in-2025
54 EU AI Act: What It Means for Agentic Commerce --- Edgar, Dunn & Company. https://www.edgardunn.com/articles/the-new-rules-for-ai-inside-the-eus-bold-legislation
55 EU AI Act 2026 Updates: Compliance Requirements and Business Risks --- Legal Nodes. https://www.legalnodes.com/article/eu-ai-act-2026-updates-compliance-requirements-and-business-risks
56 What The EU AI Framework And UK’s Approach Mean For Your Agreements --- Docusign. https://www.docusign.com/en-gb/blog/what-the-eu-ai-framework-and-uks-approach-mean-for-your-agreements
57 EU AI Act: first regulation on artificial intelligence --- European Parliament. https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence
58 How to use agentic AI in line with the EU AI Act --- CX Network. https://www.cxnetwork.com/artificial-intelligence/articles/how-to-use-agentic-ai-in-line-with-the-eu-ai-act
59 Agentic AI, Risk and Compliance Under the EU AI Act --- CMS.law (Sweden). https://cms.law/en/swe/legal-updates/agentic-ai-and-the-eu-ai-act2
60 High-risk AI in the European Union --- DLA Piper Intelligence. https://intelligence.dlapiper.com/artificial-intelligence/?t=06-high-risk-uses&c=EU
61 Agentic AI, Risk and Compliance Under the EU AI Act --- CMS.law (Austria). https://cms.law/en/aut/legal-updates/agentic-ai-and-the-eu-ai-act2
62 Red Lines under the EU AI Act --- FPF. https://fpf.org/blog/red-lines-under-the-eu-ai-act-understanding-prohibited-ai-practices-and-their-interplay-with-the-gdpr-dsa/
63 AI laws in 2026: the current EU--US landscape --- Blocshop. https://www.blocshop.io/blog/ai-laws-in-2026-the-current-euus-landscape-and-how-it-shapes-software-development
64 Annex III: High-Risk AI Systems --- EU AI Act. https://artificialintelligenceact.eu/annex/3/
65 Article 16: Obligations of Providers of High-Risk AI Systems --- EU AI Act. https://artificialintelligenceact.eu/article/16/
66 EU AI Act vs GDPR: How They Work Together --- GuruSup. https://gurusup.com/blog/eu-ai-act-vs-gdpr
67 Agentic law in the European Union: Governing autonomous AI agents --- Jurisconsul. https://www.jurisconsul.com/post/agentic-law-in-the-european-union-governing-autonomous-ai-agents
68 2026: the year AI grows up --- Taylor Wessing. https://www.taylorwessing.com/de/interface/2025/predictions-2026/2026-the-year-ai-grows-up
69 The EU AI Act: What Businesses Need To Know --- Skadden. https://www.skadden.com/insights/publications/2024/06/quarterly-insights/the-eu-ai-act-what-businesses-need-to-know
70 The Agentic AI Revolution: Managing Legal Risks --- Squire Patton Boggs. https://www.squirepattonboggs.com/insights/publications/the-agentic-ai-revolution-managing-legal-risks/
71 EU AI Act and GDPR: Tracing CJEU case law on automated processing --- Covington. https://www.cov.com/-/media/files/corporate/publications/2025/10/eu-ai-act-and-gdpr-tracing-cjeu-case-law-on-automated-processing-and-decision-making.pdf
72 Artificial intelligence --- UK Regulatory Outlook February 2026 --- Osborne Clarke. https://www.osborneclarke.com/insights/regulatory-outlook-february-2026-artificial-intelligence
73 CJEU Clarifies GDPR Rights on Automated Decision-Making and Trade Secrets. https://www.insideprivacy.com/gdpr/cjeu-clarifies-gdpr-rights-on-automated-decision-making-and-trade-secrets/
74 Consumers’ right to explanation under AI decision making --- BEUC. https://www.beuc.eu/sites/default/files/publications/BEUC-X-2026-003\_Consumers\_right\_to\_explanation\_under\_AI\_decision\_making.pdf
75 Amazon Blocks Perplexity AI Scraping: Key Implications --- i10X. https://i10x.ai/news/amazon-perplexity-ai-scraping-lawsuit
76 Cloudflare launches AI Audit and blocks unauthorized AI bots --- Cloudflare Blog. https://blog.cloudflare.com/ai-audit/
77 Web scraping after hiQ v. LinkedIn: the legal landscape --- Perkins Coie. https://www.perkinscoie.com/en/news-insights/web-scraping-legal-landscape.html
78 Amazon’s lawsuit against Perplexity rattles AI-driven search --- Hotel News Resource. https://www.hotelnewsresource.com/article139310.html
79 Google’s Universal Commerce Protocol and the Race to Wire Agentic Shopping --- Opus Research. https://opusresearch.net/2026/01/15/googles-universal-commerce-protocol-and-the-race-to-wire-agentic-shopping/
80 Agentic commerce AI tools, protocol for retailers, platforms --- Google Blog. https://blog.google/products/ads-commerce/agentic-commerce-ai-tools-protocol-retailers-platforms/
81 Walmart Partners with Google To Pioneer Agent-Led Commerce --- RetailWire. https://retailwire.com/discussion/walmart-google-agentic-commerce/
82 Agentic AI poses new challenges around online payments --- Pinsent Masons. https://www.pinsentmasons.com/out-law/news/agentic-ai-challenges-online-payments
83 Should Amazon Be Banning AI Shopping Agents From its Platform? --- RetailWire. https://retailwire.com/discussion/should-amazon-banning-ai-shopping-agents/
84 Amazon Injunction Could Change the Future of Agentic Commerce --- PYMNTS.com. https://www.pymnts.com/amazon/2026/amazon-injunction-could-change-the-future-of-agentic-commerce/
85 Watchdog Issues Warning About Letting AI Run Your Life --- Futurism. https://futurism.com/artificial-intelligence/ai-agent-manipulation
86 EU: BEUC published position paper on consumers’ right to explanation --- DataGuidance. https://www.dataguidance.com/news/eu-beuc-published-position-paper-consumers-right
87 Consumers’ right to explanation under AI decision making --- BEUC. https://www.beuc.eu/position-paper/consumers-right-explanation-under-ai-decision-making
88 AI shopping agents: How will UK Consumer Law apply? --- CMS. https://cms.law/en/gbr/legal-updates/ai-shopping-agents-how-will-uk-consumer-law-apply
89 Complying With the DSA, AI Act, and GDPR --- Goodwin Law. https://www.goodwinlaw.com/en/insights/publications/2025/07/insights-practices-antc-three-laws-one-challenge
90 When Bots Set Prices: CMA Highlights Real World Risks of Algorithmic Pricing --- Akin Gump. https://www.akingump.com/en/insights/alerts/when-bots-set-prices-cma-highlights-real-world-risks-of-algorithmic-pricing
91 AI and collusion: frontiers, opportunities and challenges --- CMA. https://competitionandmarkets.blog.gov.uk/2026/03/04/ai-and-collusion-frontiers-opportunities-and-challenges/
92 EU Regulations Are Not Ready for Multi-Agent AI Incidents --- TechPolicy.Press. https://www.techpolicy.press/eu-regulations-are-not-ready-for-multiagent-ai-incidents/
93 US Justice Dept settles Agri Stats case --- Reuters. https://www.reuters.com/world/us-justice-dept-settles-agri-stats-case-2026-05-07/
94 UKJT Public Consultation: Liability for AI Harms under the Private Law of England and Wales --- LawtechUK. https://lawtechuk.io/ukjt/public-consultation-liability-for-ai-harms-under-the-private-law-of-england-and-wales/
95 2026 AI Legal Forecast: From Innovation to Compliance --- Baker Donelson. https://www.bakerdonelson.com/2026-ai-legal-forecast-from-innovation-to-compliance
96 Could agents be the next stumbling block for Europe’s AI rules? --- Euractiv. https://www.euractiv.com/news/could-agents-be-the-next-stumbling-block-for-europes-ai-rules/
97 Driving compliance with EU’s AI Act through Agentic AI agents --- Consultancy.eu. https://www.consultancy.eu/news/12432/driving-compliance-with-eus-ai-act-through-agentic-ai-agents
98 Agentic AI: The liability gap your contracts may not cover --- Clifford Chance. https://www.cliffordchance.com/insights/resources/blogs/talking-tech/en/articles/2026/02/agentic-ai-and-the-liability-gap-your-contracts-may-not-cover.html