February Insights

By Eve, Artstash Creative · Published 25 February 2026 · Last updated August 2026

Quick answer: February was dominated by the gap between what AI can now produce and what teams should do with it. ByteDance released Seedance 2.0, capable of video realistic enough to draw a cease and desist from Disney. Tobias Charles argued in a widely read piece that AI is making managers worse by letting them skip the thinking. And both Gartner and Adobe published data pointing the same way: search volume and referral traffic are shifting toward AI surfaces. The through-line is that AI rewards teams who already know what they are trying to say.

What Is ByteDance Seedance 2.0?

ByteDance recently released its new Seedance 2.0 model, allowing for stunningly realistic video generation. Capable of rendering up to 12 seconds at full 1080p resolution, and at a fraction of the cost of competing models, the tool shows enormous potential for becoming a mainstay in any marketer’s toolbox. The release seems impressively ahead of its time, leading to much speculation about how competitors like Sora and Kling will respond themselves.

The generated videos bear such an uncanny realism that ByteDance has been struck with a Cease and Desist from the Walt Disney Corporation over allegedly using their characters to train the model. If you are looking to use Seedance’s new 2.0 model for your marketing needs, do be sure to make your output steer clear of any infringement. Our CEO commented that actors and public figures that are so inclined might deem it wise to rent or lease out rights to their likeness for films, much in the same way OpenAI and Disney found a similar agreement.

How Is AI Making Managers Worse?

Last month Tobias Charles wrote a Medium article titled “How AI Is Making Managers Worse”. It described how many managers began relying on tools like ChatGPT to review their reports and generate insights from conversations, but that this deferral comes at the cost of authenticity and accuracy:

AI feels easy to talk to. It listens. It answers quickly. It rarely pushes back. But human relationships don’t work like that. Trust needs friction. If a person never says ‘no’, how can you trust their ‘yes’?

Charles goes on to explain their answers to this trap:

Avoid scripted feedback and start with your own ideas. Write out your thoughts before writing prompts. This will help avoid your creative muscles atrophying.

Guard against fake authority and check facts yourself. Read original sources. Do not trust citations by default.

Protect your judgement and treat outputs as drafts. Rewrite them in your own words. Add real detail from your team.

Be transparent with your use of it. Hidden dependence erodes trust (and most people can guess when you’re using it anyway).

Use a few systems to cross-check each other.

Watch your emotional response. Relief and validation signal risk: that you’re seeking confirmation. As a leader, always be wary of ‘yes men’. And AI is the pinnacle of this. Instead, test your ideas with someone who disagrees, and find lessons in the friction.

In summary: AI is a tool. And like any tool, it can only serve its wielder if used for its intended purposes. Misuse of the tool can only hinder the parties involved – this is further complicated by the fact that blind faith in AI tools can damage your critical thinking and patience without you even realizing.

What Are the Big AI Shifts for Creative Teams?

Our Creative Director has been keeping an eye on how the industry has been using AI, and these were his major findings this month:

Gartner: Search Volume Shifting Toward AI

Google’s AI Overviews provide information sourced from various sites across the web without the need to leave the search results page. Because of this, AI Overview and similar AI agents are slowly becoming big players in terms of search volume: More and more search queries are being satisfied without increasing traffic to individual websites. Gartner predicts that this will be the norm, with Vice President Analyst Alan Antin stating that this shift will “force companies to rethink their marketing channels strategy as GenAI becomes more embedded across all aspects of the enterprise.”

Gartner predicts that search engine volume will drop by a quarter in the coming months, with AI agents picking up that market share – And it’s already starting to prove true.

Adobe: AI-Driven Traffic Surge

Adobe Analytics’ recently released data review states that, since last year, sites have seen a massive spike in traffic referred from AI agents. The main contributing factor to this trend was the end of year holiday shopping season, where retail sites found the biggest traffic increase of 693%.

Vivek Pandya of Adobe for Business called this spike “a turning point for how consumers discover and choose brands” But while AI is helping steer consumers towards certain brands, the part it plays in marketing those brands may be changing as well.

Summary – The shift away from AI-generated creative

We’ve noticed that AI-generated ads are becoming more and more scarce in recent weeks. There was a bubble experimenting with all these new tools as they came out, but it seems they are being used more sparingly and subtly. However, using AI in production has really picked up on the backend, to enhance and speed up production pipelines. Subtlety in the use of AI might be the key to winning brand loyalty conversion, as it appears that people are becoming less and less trustworthy of brands using AI for assets. As the articles above mention, it seems like AI is shifting towards being used heavily in marketing analytics for campaigns.

But all these headlines confirm what sharp-eyed experts have already suspected: AI is a tool, and it only facilitates life as long as it is used effectively. The question becomes, how are you using it?

Works Cited

Bajarin, Tim. “How AI Is Rewriting the Rules of Marketing.” Forbes, 20 Jan. 2026, www.forbes.com/sites/timbajarin/2026/01/20/how-ai-is-rewriting-the-rules-of-marketing/.

Charles, Tobias. “How AI Is Making Managers Worse.” Medium.com, 9 Jan. 2026, medium.com/psychology-of-workplaces/how-ai-is-making-managers-worse-df9f4cb09eec. Accessed 16 Feb. 2026.

Gartner. “Gartner Predicts Search Engine Volume Will Drop 25% by 2026, due to AI Chatbots and Other Virtual Agents.” Gartner, 2024, www.gartner.com/en/newsroom/press-releases/2024-02-19-gartner-predicts-search-engine-volume-will-drop-25-percent-by-2026-due-to-ai-chatbots-and-other-virtual-agents.

McCoy, Kimeko. “With AI Backlash Building, Marketers Reconsider Their Approach.” Digiday, 12 Feb. 2026, digiday.com/marketing/with-ai-backlash-building-marketers-reconsider-their-approach/.

Pandya, Vivek. “AI Traffic Surges across Industries, Retail Sees Biggest Gains.” Adobe.com, 2026, business.adobe.com/blog/ai-driven-traffic-surges-across-industries.

Seedance. “Seedance AI.” Seedance AI, 2026, www.seedance.ai. Accessed 16 Feb. 2026.

Frequently asked questions

What is Seedance 2.0?

ByteDance’s video generation model, capable of producing footage realistic enough that the company received a cease and desist from The Walt Disney Company over generated likenesses. It marks the point where generated video realism became a legal problem rather than a technical one.

How is AI making managers worse?

Tobias Charles’ argument is that AI lets managers skip the reasoning work that used to produce judgement. Delegating analysis to a model without engaging with the problem produces plausible-sounding decisions that nobody on the team can interrogate or defend.

Is search traffic really shifting to AI?

Gartner projects meaningful decline in traditional search volume as AI answer surfaces absorb informational queries, and Adobe has reported a surge in AI-driven referral traffic. Both point the same direction: fewer clicks from classic results, more from AI answers.

What should creative teams do about AI answer engines?

Make content extractable. Clear direct answers near the top, question-shaped headings, and structured data give AI systems something clean to lift and attribute. Volume of content matters far less than whether a machine can parse a straight answer out of it.