CONSUMER

With new CEO, Apple picks a lane in the AI race

Apple is heading into the AI era with a hardware guy as its new CEO. 

Longtime hardware engineering leader John Ternus will succeed Tim Cook as the company's new chief executive on September 1, the company announced on Monday. 

While Steve Jobs turned Apple into the top consumer product company in the world, Tim Cook took that success and turned it into the most profitable consumer business of the era and the world's most valuable company for a long stretch from 2011 to 2024. 

But its next act is a much bigger question mark. 

In the past two years, Apple was dethroned from its spot as the No. 1 public company as investors poured money into Nvidia and Google, two leaders of the AI boom. Both companies passed Apple, while Microsoft and Amazon have also bet their futures on AI and threaten to overtake Apple in the months and years ahead, if it can't find its place in the AI ecosystem. 

When it comes to Apple and AI, there are a number of conflicting trends to follow:

• R&D: Apple has regained some investor confidence lately because of its more sober approach to AI R&D spending compared to the other tech giants. While that works in its favor in a risk-off market, if and when investors get bullish about AI again, Apple will get left behind for the same reason. Investing less in future projects could also limit Apple's longterm possibilities for its next major product hit.

• Gemini: Earlier this year, Apple waved the white flag on becoming a frontier AI lab and signed a deal with Google to white-label Gemini as Apple's AI model provider. Primarily, Gemini will give Siri a brain transplant in the next version of iOS. 

• Agents: The personal AI agent boom has turned into an unexpected win for Apple, but not because of the software. AI enthusiasts have rushed to buy Mac mini and Mac Studio computers to run their agents in a safe, separate box. The boom has turned the Mac mini into a bit of a cult hit and an icon of the AI agent moment of 2026. In fact, sales have been so brisk that both the Mac mini and Mac Studio are backordered until the fall.

• Devices: In recent years, Apple also made the wrong bet on VR headsets with Vision Pro, rather than focusing on lightweight AR experiences with glasses—a form factor that's also much better suited for the AI future. It's now retrenching and reportedly preparing to launch not only glasses but also other AI-first devices. 

"Ternus represents a quiet pivot back toward product intimacy, a tighter coupling between hardware, software, and emerging AI capabilities," said Dipanjan Chatterjee, principal analyst at Forrester. "But he must resist the temptation of incrementalism that has plagued Apple of late and escape the iPhone’s gravitational pull in his quest for the next disruptive form factor. As Ternus assumes the helm, he must define Apple's future as ferociously as he defends its past."

Ternus is a hardware product leader through and through, and the timing of this transition is not accidental. Apple has reached crunch time in its AI journey, and it needs a clear strategy. There's still time to play a key role in AI's future, but it has catching up to do, and the margin for error has decreased considerably. With Ternus taking the reins, Apple is again making the statement that it sees its role in the AI ecosystem primarily as a hardware and devices builder. That means it will have to win on integration, ease of use, privacy, and trust. While it has the home-field advantage in all those areas, it will have to avoid the unforced errors that have plagued its AI execution over the past two years.

TOGETHER WITH MIRO

The Secret Sauce To Nailing Product Development

If you’ve ever been involved in bringing a product to market, you know a few things: 

1. It’s a highly strategic process grounded in solid research, design, and planning, and…

2. It rarely goes according to plan.

That’s exactly why Miro has put together this free webinar on how to improve your product development process with one simple trick: Early, rapid, and collaborative prototyping.

In this webinar, you’ll learn how developing and sharing a prototype sooner than you think can help align all your teams’ visions, iron out roadbumps, and get every stakeholder on the same page before your product goes into coding. It’s completely free, so register for your spot right here before it fills up.

ENTERPRISE

Adobe thinks it has the missing layer for agents

The theme of Adobe’s flagship customer experience conference? You guessed it: agents. It's still 2026, after all. 

On Monday at Adobe Summit, the company launched a slew of agentic products for marketing professionals, with the most notable announcement being Adobe CX Enterprise, an agentic AI system that businesses can use to create agentic workflows to manage the entire customer lifecycle, according to the blog post. 

“Adobe CX Enterprise enables businesses to scale agentic AI with a fully customizable solution that is tailored to the needs of their organization, moving teams beyond AI experiments to tangible business outcomes,” said Anil Chakravarthy, President, Customer Experience Orchestration Business at Adobe, in the release. 

In practice, Adobe CX Enterprise is meant to make it easier for organizations to deploy agentic systems that deliver personalized experiences aligned with brand guidelines while also adhering to governance principles and protocols, such as MCP. The platform features: 

• New agents: These agents can streamline tasks across Adobe applications and third-party ecosystems, all easily accessible in one hub. 

• Agent skills catalog: These “skills” or  reusable instructions make it easier for businesses to deploy repeatable workflows

• Adobe CX Enterprise Coworker: This tool is designed to help coordinate and execute AI agents to ensure they meet business objectives through a conversational interface. It will be generally available in the upcoming months, but Adobe didn't give specifics. 

On the Summit Stage, a demoer, acting as a VP of Marketing at Marriott Bonvoy, used Adobe CX Enterprise to evaluate a potential customer acquisition and engagement opportunity from the South of France, suggest a course of action, and include who to reach out to and complete the assignment. 

Meanwhile, the colleague she reached out to was able to use CX Enterprise to not only surface the request but also help understand why the conversion rate was so low for travel bookings in the given area despite the traffic spike, get suggestions on how to fix the page to address the issues, and even have agents automatically redesign the landing page.

Adobe CX Enterprise is open and interoperable, allowing maximum flexibility and functionality with partners such as AWS, Anthropic, IBM, OpenAI, and more. Adobe customers already use the Adobe Experience Platform to bring together diverse data sources for insights and orchestration, and now that same platform serves as the contextual layer for the new CX Enterprise experience.

While AI agents generated a lot of buzz in the enterprise in 2025, we're still hearing so much about them a year later because many of those initiatives never fully left the pilot phase. For AI agents to be successful, it takes much more than the underlying technology, which has long been available. The challenge is finding the right ways to deploy within a business that gives it the right level of contextual data and guardrails. Adobe's CX Enterprise attempts to address both challenges by making it interoperable with other partner solutions and by making MCP and similar integrations easier to deploy.

Disclosure: Sabrina Ortiz's travel to Adobe Summit was paid by Adobe. The Deep View's coverage is editorially independent from the companies we cover.

TOGETHER WITH QA TECH

Where AI velocity breaks

Version control is automated. Deployments are automated. Code review has AI assistance.

Yet most teams still rely on a human to validate releases – or skip it entirely because there isn’t time.

The same AI accelerating your development velocity is making manual QA impossible to scale – especially as release frequency increases.

Teams using QA.tech reclaim 320+ hours every month – not by hiring more QA engineers, but by replacing manual validation with agents that automatically validate every release.

See how it works.

PRODUCTS

Sneak peek at Adobe’s next AI feature

I have attended Adobe Summit three years in a row, and the best part of the event is the "Sneaks": the company's preview of cutting-edge experiments that may one day join its suite of tools.

The submissions are crowdsourced from employees across the company, each vying for a chance to present their big ideas and potentially ship the next breakout feature, with 40% historically making it to full rollout. Some of Adobe's most popular features were first showcased at Sneaks, including Generative Fill, which was originally unveiled as Project Fast Fill.

This year, the team received over 500 submissions, up from 150 last year, and had to whittle them down to the top seven. Ahead of Sneaks this year, I got an additional sneak peek into one of the features being showcased—Project Tailored Takes—so let's spotlight that one.

The concept is simple: it takes Firefly Foundry, Adobe’s enterprise generative AI models that are customized to the company’s data and assets, and applies them to generate videos that serve marketers’ needs. This includes first creating images that reflect your needs, then using them to build storyboards, and finally generating variations tailored to different audiences. 

For instance, in a demo, the user began with an origin-story campaign for a global coffee brand. A quick prompt enabled Firefly Foundry to generate on-brand product imagery, and the user was then able to create multiple brand-consistent variants with the product or background swapped, ready to be used as potential video frames. Other tweaks included camera view adjustments, A/B Testing, and localization for different parts of the globe.

“I think the big problem that is solved is all about scale and speed, and so getting from interesting assets to valuable code or production-ready assets more quickly,” Eric Matisoff, principal evangelist at Adobe, told The Deep View.

The sheer increase in internal submissions Adobe received from employees for Sneaks this year is notable in itself. Not only are agentic solutions being built for enterprise customers and creatives, but those same AI tools are also making it easier for people to create cutting-edge features and experiments. The amount of innovation that agentic AI tools are fueling is exciting, especially because we are just scratching the surface of what these tools are enabling professionals inside of companies like Adobe to create.

Disclosure: Sabrina Ortiz's travel to Adobe Summit was paid by Adobe. The Deep View's coverage is editorially independent from the companies we cover.

LINKS

• Apple CEO Tim Cook steps down, John Ternus steps up

• Amazon invests addiitonal $25 billion in Anthropic to build out AI infrastructure

• Cloud platform Vercel confirms a security leak as hackers try to sell stolen data

• Music app Deezer claims AI-generated tracks amke 44% of daily uploads

• China’s Netflix equivalent, IQiyi, expects bulk of shows to be AI-created in next five years

• Lovable denies involvement in data breach amid concerns over chat and code visibility

• Kimi K2.6: Moonshot open-sourced its latest model 

• Qwen3.6-Max-Preview: Alibaba introduced an early preview of its next flagship model

• Claude Cowork: Users can now build live artifacts (dashboards and trackers)

• Codex: OpenAI launched “Chronicle” experimental tool

• EY: AI & Machine Learning Engineering Consultant

• Perplexity: AI Security Engineer

• Perplexity: Engineering Manager (AI Inference)

• Toast: Product Manager, AI Marketing Agent

GAMES

POLL RESULTS

In your experiments with AI agents, have they met your expectations?

Agents have blown me away (39%)
Agents have been underwhelming so far (32%)
Agents have met my expectations (29%)