Narrative Cycle Report: Consumer AI

1. The Category Snapshot

The term generating the most media coverage in consumer AI right now is not the one that dominated 18 months ago. And the term that will dominate 12 months from now is already accelerating at 188% quarter-over-quarter growth.

We tracked seven distinct narratives across 39 months of continuous media data (January 2023 through March 2026), covering over 200,000 English-language news sources. Here is where each one sits today:

The conventional view is that consumer AI is an “agentic AI” story right now, and that positioning efforts should focus there. The data partially supports this. Agentic AI has the strongest media-buyer alignment of any narrative: high coverage volume, high search demand (110,000 monthly searches), and the deepest enterprise CPC pool ($20 to $136 per click). But it also reveals something the conventional view misses: narrative cycles in this category are compressing by roughly 33% each generation, the successor cycle is already in motion, and the window to position for it is shorter than most companies assume.

2. The Cycle Data

Generative AI: The predecessor that refused to decline

Generative AI established the first complete narrative cycle in our dataset. It emerged in January 2023 at 4,681 articles, surged to a peak of 124,440 articles per quarter in Q2 2023 (driven almost entirely by ChatGPT coverage), then entered a sustained decline through Q1 2025, bottoming at 80,999 per quarter. A 35% drawdown from peak.

Then it did something the standard lifecycle model doesn't predict: it came back. Q1 2026 volume reached 138,689, exceeding the original peak by 11%. The coverage shifted from hype (“what is generative AI?”) to implementation (“how enterprises are deploying generative AI”). The term evolved from a novelty descriptor to an operational category.

Cycle timing: 6 quarters from first signal to peak surge. 8 quarters from peak to trough. Resurgence began ~10 quarters after peak.

Agentic AI: The dominant cycle, still accelerating

Agentic AI first appeared at meaningful volume in Q1 2024 (2,931 articles per quarter). It hit its inflection point in Q4 2024, when quarter-over-quarter growth reached 201.8%. By Q1 2026, it had reached 180,822 articles per quarter, with no plateau signal yet. March 2026 alone produced 98,484 articles, the single highest-volume month for any narrative in the dataset.

The acceleration was anchored by specific corporate events: Salesforce's Agentforce launch, Microsoft's Copilot agent framework, and the broader enterprise AI agent ecosystem that consolidated in late 2024.

Cycle timing: 4 quarters from first signal to peak surge. That's 33% faster than generative AI's equivalent journey.

Embodied AI: The fastest-growing narrative in the dataset

Embodied AI was essentially invisible before 2025. Quarterly volumes were below 400 through all of 2024. Then Q1 2025 hit 801, Q3 2025 reached 1,946, and Q1 2026 exploded to 6,790: a 187.7% quarter-over-quarter acceleration. March 2026 alone produced 3,877 articles.

What makes this trajectory distinctive is the presence of multiple simultaneous corporate anchors. NVIDIA has been seeding “physical AI” as a parallel term. The Chinese government designated embodied AI as a strategic national priority in 2025. And the humanoid robotics wave (Figure, Tesla Optimus, Agility Robotics) is generating coverage that increasingly folds into the “embodied AI” umbrella. At the equivalent lifecycle stage, agentic AI was growing at roughly half this rate.

AI Companion and Personal AI: Two terms, one convergence

“AI Companion” has been present in the dataset since Q1 2024, but its pattern is unusual. Rather than sustained acceleration, it follows a sawtooth: volume spikes on product events (Character.AI milestones, Meta's companion features) then retreats. Q1 2026 at 10,567 is its highest quarter, but the growth is uneven. No corporate anchor has claimed the term.

“Personal AI” follows a different trajectory. Volume is lower (5,421 in Q1 2026) but more consistent, and the Q1 2026 growth rate of 123% suggests early acceleration. More importantly, buyer intent signals for “personal AI assistant” (14,800 monthly searches, $6.45 CPC) are forming ahead of media coverage. In most narratives, media leads buyers. Here, buyers are leading media.

These two terms are converging toward the same narrative space: AI that is individually yours, that knows your context, that adapts to you. “Personal AI” is the broader frame. It can absorb “companion” (consumer) while extending into enterprise use cases.

The Compression Finding

The most operationally significant pattern in the data: each successive dominant narrative reaches peak surge faster than its predecessor.

If this pattern holds, the next dominant cycle will reach peak surge in 2 to 3 quarters from its inflection point. For companies trying to time their positioning, this means the window between “this narrative is emerging” and “this narrative is peaking” is shrinking from 18 months to under 12.

3. The Buyer Map

Media coverage tells you what the press is writing about. It does not tell you where buyers are spending. We cross-referenced the narrative volume data against search demand and cost-per-click (CPC) data across 40 keywords to measure actual commercial intent.

Where media and money are aligned

Agentic AI has the strongest media-buyer alignment in the dataset. The “AI agent SDK” keyword at $136.19 per click signals enterprises in active procurement mode. Generative AI remains aligned but more competitive; lower CPC ($11.40) indicates a saturated, mature buyer market.

Where media runs ahead of buyers

Embodied AI has growing media attention but thin buyer pools. Search volume for “embodied AI” is 1,900/month at $11.25 CPC. The CPC exists (some enterprises are bidding) but the volume is thin. Applying the 1\u20132 quarter media-to-buyer lag, buyer pool formation signals should appear by Q2\u2013Q3 2026.

Where buyers run ahead of media

“Personal AI assistant” generates 14,800 monthly searches at $6.45 CPC, but “personal AI” as a media narrative only produced 5,421 articles in Q1 2026. Buyers are looking for something the press hasn't fully named yet.

Where media and money are misaligned

AI Companion generates meaningful media coverage (10,567 articles) and meaningful search volume (14,800/month) but CPC is $2.67. For comparison, agentic AI keywords range from $20 to $136. Low CPC means consumer browsing, not enterprise evaluation.

The media-to-buyer lag

In this category, the lag between media narrative acceleration and buyer pool formation is 1 to 2 quarters. This is unusually short compared to prior technology cycles (cloud computing showed a 4 to 6 quarter lag). When you see a narrative accelerating in media, the buyer window opens within 3 to 6 months, not 12 to 18.

4. The Overlooked Signal: The Resurgence Paradox

The most counterintuitive finding in this dataset is a number: 138,689.

That's the quarterly article count for “generative AI” in Q1 2026. It exceeds the original peak of 124,440 in Q2 2023 by 11%.

This shouldn't happen. The standard narrative lifecycle model, and the Gartner Hype Cycle that most of the market uses as a mental model, assumes each cycle declines permanently once its successor emerges. Generative AI was supposed to be the term that peaked in 2023, faded through 2024, and settled into background noise as “agentic AI” took over. Instead, it came back stronger.

What's happening is a term-level evolution. “Generative AI” in 2023 was a novelty label attached to ChatGPT and wonder. “Generative AI” in 2026 is an enterprise deployment category. The coverage shifted from “what is this?” to “how is it being implemented?” The same words now carry different editorial weight.

This pattern is invisible to snapshot analysis. It only becomes visible when you lay the curve against 39 months of continuous data and the lifecycle assumption breaks down.

The implications

First, the total addressable narrative volume in consumer AI is expanding, not rotating. Each new narrative adds to the total rather than stealing share from its predecessors.

Second, companies don't need to “get off” a narrative before it peaks. The fear that being associated with “generative AI” makes you look dated is contradicted by the data. The term still commands 138,000+ articles per quarter.

Third, the portfolio approach outperforms the serial approach. Companies that are credible across multiple narrative layers simultaneously have more surface area for media coverage and buyer engagement than companies that ride one wave and try to jump to the next.

5. The Forecast

Based on 39 months of cycle data, compression patterns, and buyer intent signals, here is where we see consumer AI narratives heading over the next 12 to 18 months.

Embodied AI: the next dominant narrative (75% confidence)

Embodied AI is following the same trajectory as agentic AI, but faster. At the equivalent lifecycle stage, agentic AI was growing at half the rate. Applying the 33% compression pattern, peak surge arrives in 2 to 3 quarters from the current inflection: late 2026 to early 2027. The narrative has structural advantages: multiple simultaneous corporate anchors (NVIDIA, Chinese government, humanoid robotics companies), a logical semantic progression (generative to agentic to embodied), and active sub-narrative absorption.

What could change this: a major robotics safety incident, regulatory action on autonomous systems, or NVIDIA rebranding away from “physical AI.” None are likely in the near term.

Personal AI: the parallel consumer narrative (60% confidence)

“Personal AI” and “AI Companion” are converging, and “personal AI” is the broader, more durable label. Buyer intent data shows real demand forming (14,800 monthly searches) ahead of media attention. This narrative will not replace agentic or embodied in business press. It will grow in parallel across consumer, lifestyle, and culture beats. The ignition event needed: a consumer AI product reaching 100M+ users under the “personal AI” label.

Multimodal AI: absorption, not ascension (30% confidence)

“Multimodal AI” is growing but decelerating (7.0% QoQ in Q1 2026, down from 47.7% a year earlier). No corporate anchor has claimed the term exclusively. It is being absorbed as a capability descriptor (“multimodal agents,” “multimodal models”), following the pattern of “cloud-native” in the 2010s.

6. The Positioning Windows

Three distinct positioning windows open in the next 12 to 18 months. Each suits a different company profile.

Window 1: The Embodied AI Bridge (Now through Q3 2026)

Window 2: The Personal AI Thought Leadership Play (Now through H1 2027)

Window 3: The Portfolio Position (Ongoing)

The difference between companies that ride narrative cycles and companies that get caught in them comes down to one thing: whether they're watching continuous data or reacting to what everyone else is already talking about. By the time a narrative feels obvious, the positioning window is already closing.