78% Of Marketers Are Now Using Generative AI: How Veo 3, Seedance 2.0, Sora 2 And Brand Guardrails Are Reshaping UK Advertising In 2026

78% — Marketing teams now using generative AI in at least one stage of ad production — up from 41% in 2024  ·  $190B+ — Projected global AI ad spend in 2026 — the structural shift from human-only to AI-augmented creative production  ·  70 million — Gemini-generated creative assets inside Google's Performance Max and AI Max campaigns in Q4 2025 alone — 3x year-over-year  ·  Sora 2 / Veo 3 / Gen-4 / Seedance 2.0 — The four native-video models now embedded directly inside the major ad tool stacks

The generative AI advertising shift has happened. 78% of marketing teams now use generative AI in at least one stage of ad production — up from 41% in 2024. Global AI ad spend is projected to surpass $190 billion in 2026. Native video models (Sora 2 from OpenAI, Veo 3 from Google, Runway Gen-4, ByteDance's Seedance 2.0) are now embedded directly inside the major ad tool stacks rather than being separate creative tools. Agentic workflows — where AI plans, generates, ships and iterates on campaigns autonomously — have become the new baseline for UK CMOs. Google reported that advertisers generated nearly 70 million creative assets inside Performance Max and AI Max campaigns in Q4 2025 alone, a 3x year-over-year increase. Brand Guardrails technology now prevents AI from generating content that deviates from a brand's hex codes or logo usage policies, addressing the brand-safety concern that previously slowed deployment.

For UK CMOs and brand marketers, the 2024-era playbook — human-only creative concepting, manual asset production, sequential review cycles, separate creative tools versus media-buy tools — is structurally obsolete. The 2026 playbook is genuinely different. AI handles the bulk of asset variant generation. Native video models live inside the media-buying interface. Agentic campaign workflows operate autonomously within governance guardrails. Brand safety is enforced by AI-aware policy systems rather than manual review. The competitive pressure on UK brands to operate at the new baseline is real — competitors using gen-AI advertising tools at scale are launching more campaign variants, iterating faster on early signals, localising more efficiently across markets, and capturing meaningfully better unit economics on paid media. This is the complete UK marketer's guide to operating in the gen-AI advertising era — the tooling, the agentic workflows, the brand-safety architecture, the ASA compliance landscape, and the 90-day H2 2026 rollout plan.

The Four Native Video Models UK Marketers Should Understand

Sora 2 (OpenAI)

OpenAI's Sora 2 is the strongest general-purpose AI video generation model for narrative quality, scene coherence, and physical realism. For UK brands producing premium brand-led video where production values matter most, Sora 2 is typically the right choice. It is also the most-integrated with the broader OpenAI ecosystem, meaning teams already using ChatGPT for ideation and copy can hand off to Sora 2 for video generation without leaving the workflow.

Veo 3 (Google)

Google's Veo 3 is the highest-volume production AI video model in the market thanks to its native embedding in Performance Max and AI Max — Veo 3 is the model behind a substantial share of the 70 million Q4 2025 creative assets Google generated for advertisers. For UK brands operating heavy Google paid-media programmes, Veo 3 is structurally the right choice because it sits inside the media-buying interface rather than requiring separate workflow. Expect substantial Veo 3 announcements at Google I/O 2026 tomorrow.

Runway Gen-4

Runway Gen-4 is the choice of creative-led production teams and agencies that prioritise creative control and editing flexibility. Runway has built deep integrations with Adobe Premiere, CapCut, and other professional editing tools, making Gen-4 the typical choice for UK creative agencies producing AI-augmented commercial work where the creative team needs fine-grained control over output.

Seedance 2.0 (ByteDance)

ByteDance's Seedance 2.0, announced in February 2026, is the AI video model that genuinely surprised the market. Its defining capability is producing coherent multi-shot commercial arcs from a text prompt or a single product photo — a full commercial sequence from product reveal to lifestyle shot to call-to-action, ready for paid social deployment. The launch went viral in China and has rapidly been adopted by UK direct-to-consumer brands operating heavy short-form social paid-media programmes (TikTok, Instagram Reels, YouTube Shorts). For UK D2C brands, Seedance 2.0 is increasingly the default choice for short-form social ad creative.

Agentic Advertising Workflows — The New Baseline

The structural shift in 2026 is that gen-AI capability has moved from 'tool you use to make individual assets' to 'agentic workflow that plans, generates, ships and iterates on entire campaigns autonomously within defined guardrails'. The pattern works as follows. The CMO or campaign manager defines campaign objectives, target audience, brand guardrails, budget envelope and success metrics. The agentic platform (typically Performance Max with AI Max for Google-led campaigns, Advantage+ for Meta-led, plus a long tail of independent platforms) generates the initial creative-asset variants, places the initial media buy, monitors early performance signals, generates additional variants based on what is working, and iterates the campaign continuously across its lifecycle.

For UK marketing teams, the implication is a substantial reduction in the volume of routine campaign-execution work and a substantial increase in the importance of campaign-design work — defining the brief, setting the guardrails, choosing the success metrics, and overseeing the agentic execution. The 2024-era media-buying team that spent its time placing buys, generating variants and adjusting bids is being progressively automated. The 2026-era marketing team spends its time on strategic campaign design, creative-direction quality, brand-safety governance, and the structured measurement of agentic campaign output. UK CMOs that are not redesigning their marketing organisation for this shift are over-staffed for the agentic-execution layer and under-staffed for the strategic-design layer.

The UK ASA Position Marketers Cannot Get Wrong

The UK Advertising Standards Authority's position on AI-generated advertising content tightened materially through Q1 and Q2 2026. The ASA now requires clear disclosure when content is materially AI-generated, when AI tools are used in production beyond minor editing, when virtual influencers or AI-augmented human influencers are used in brand-paid content, and (increasingly) when AI is used to personalise creative dynamically at the individual user level in ways that could be seen as manipulative. Non-compliance has resulted in formal ASA rulings against several UK consumer brands through Q1-Q2 2026 with reputational consequences materially larger than the formal penalty implies.

UK marketing leaders deploying agentic advertising workflows must design compliance into the workflow from day one rather than bolting it on later. The Brand Guardrails layer is the natural place to enforce ASA-compliant disclosure as a structural feature — every AI-generated asset gets the disclosure treatment built in. Marketing teams that defer compliance design end up either retrofitting at higher cost or accepting genuine regulatory and reputational risk. The window to design this in elegantly closes as the campaign portfolio scales.

The 90-Day UK CMO Generative AI Advertising Rollout Plan

1. Days 1-14: Audit your current marketing tech stack and creative workflow. Map where gen-AI is already being used (often more than the CMO realises), where it is not being used but should be, and where Brand Guardrails are currently absent. The audit is the foundation for any meaningful AI-augmentation decision.
2. Days 15-30: Pick your primary agentic-platform partner for H2 2026. For most UK brands this is Performance Max + AI Max (Google) for Google paid media and Advantage+ (Meta) for Meta paid media, with Seedance 2.0 and Veo 3 for video creative generation. Build the structured success-measurement framework against current-baseline campaign performance.
3. Days 31-50: Deploy Brand Guardrails as a structural platform feature. Encode hex codes, logo usage policies, brand tone profiles, content boundaries and ASA disclosure requirements into the agentic-platform configuration. This is the load-bearing brand-safety investment.
4. Days 51-70: Restructure the marketing team for agentic execution. Reduce the team time spent on routine campaign-execution work and reallocate to strategic campaign design, creative direction quality, brand-safety governance, and structured agentic-campaign measurement.
5. Days 71-90: Build the ASA-aligned compliance framework end-to-end. Disclosure as a structural feature, virtual-influencer policy (covered in Batch 13's creator economy piece), copyright and IP clarity, audit trail for every AI-generated paid asset. Compliance integration is what enables sustainable scaling.

Sources

1. StackAdapt — AI In Advertising: How It's Transforming Marketing In 2026
2. Marketing Brew — How Meta's AI Push Is Changing Ad Creation
3. Ad Age — Top 5 AI Marketing Activations: The RealReal, Meta, Coachella
4. Cometly — 9 Best AI Ad Creative Generators: Complete 2026 Guide
5. Chat-Data — The 5 Best AI Ad Generators In 2026 For Effortless Advertising
6. JumpFly — AI In Online Advertising: 5 Key Trends From February 2026
7. Uplifted — 12 Best AI Tools For Creatives 2026: Ship Winning Ads Faster
8. Resource Digen — 10 Best Generative AI Video Tools For Marketing In 2026
9. Smart Magic Productions — How AI Predicts Viral Ad Concepts Before Production
10. Google Ads Blog — Performance Max And AI Max Q4 2025 Creative Asset Volume
11. ByteDance — Seedance 2.0 Launch Documentation
12. UK Advertising Standards Authority (ASA) — Guidance On AI-Generated Content Disclosure
13. BraivIQ — Batch 13 AI Creator Economy Article (Internal Reference)