AstraZeneca is deploying 5,000 M5 iPad Pros to its pharma sales teams. Snowflake has put 9,000 Macs into the hands of its workforce. Copel, the Mexican energy utility, is rolling out more than 10,000 iPads across its retail operations. By any reasonable measure, enterprise has voted: Apple devices remain the hardware platform of choice.
Apple’s Q1 FY2026 earnings only reinforced that momentum. The company posted record revenue of $143.8 billion, up 16 percent year on year, while its installed base climbed past 2.5 billion active devices. On the earnings call, CFO Kevan Parekh put the enterprise trend plainly:
“Organizations are continuing to expand their fleet of Apple devices to drive productivity.”
That sounds like a clean enterprise success story. It is. But there is a more uncomfortable question underneath it: Apple’s hardware bet is working – but is its AI bet? As enterprise software shifts from assistant-style productivity features toward agentic automation, the gap between where Apple’s AI capabilities are today and where enterprise buyers may need them tomorrow is becoming harder to ignore.
Related Articles
- Is Your Automation Strategy Missing the Orchestration Layer?
- Forrester: Why Your AI Tools Might Be Making Things Worse
- Are AI Copilots Failing to Deliver Real Productivity?
The Market Is Moving Fast – and Apple Is Not Setting the Pace
Agentic AI is quickly becoming the real prize in enterprise productivity. The conversation has already moved beyond chat assistants that summarise, suggest, or retrieve information. The next phase is about systems that can take multi-step action across software, execute workflows, and operate applications with much less human prompting.
The market data makes clear how seriously organisations are taking that shift. According to Fortune Business Insights, the global agentic AI market is projected to grow from $7.29 billion in 2025 to $139.19 billion by 2034. Meanwhile, Cognipeer reports that by 2030, 45 percent of organisations expect to be orchestrating AI agents at scale across core business functions.
OpenAI is building what it describes as an agentic “super app”, combining ChatGPT, Codex, and browser-based computer use into one service capable of taking action on a user’s behalf. Anthropic is pushing Claude deeper into enterprise workflows through Claude Code and Claude for Work. These are not simply improved chatbots. They are AI systems designed to operate software, complete tasks, and manage processes.
Apple Intelligence, by contrast, is still trying to prove it can deliver a reliable assistant. Alex Kantrowitz, host of the Big Technology Podcast, summed up the problem bluntly:
“Apple still hasn’t gotten chatbot right; and OpenAI and Anthropic are already on to the next thing.”
For enterprise IT leaders planning a three-to-five-year digital workplace strategy, that is not a theoretical problem. If agentic AI becomes the next real productivity layer – and most signals suggest it will – then Apple’s current AI trajectory could leave a structural hole in the stack just as enterprises need it filled.
Siri’s Credibility Problem Is Also an Enterprise Problem
The gap between Apple’s AI ambition and its AI delivery is not new. What has changed is the cost of that gap. Siri launched in 2011, yet in April 2026 it still is not competitive with the tools many enterprise employees are already using on their own.
“Here we are in April 2026, nearly four years after ChatGPT launched – and the thing is still terrible.”
Apple is reportedly testing a standalone Siri app, according to Bloomberg’s Mark Gurman, with persistent conversation history, pinned chats, cross-session search, and Dynamic Island integration on iPhone. That would matter. One of Siri’s biggest weaknesses is still its lack of memory and continuity.
On the Q1 FY2026 earnings call, Apple confirmed that a “more personalised Siri” is coming “later in 2026”. But the context is awkward. The company had to pull promised Apple Intelligence features after WWDC 2024 because they were not ready. WWDC 2026, exactly two years later, now looks less like another product showcase and more like a credibility test.
For enterprise buyers, that goes beyond product polish. Employees already using ChatGPT, Gemini, or Claude for real work will not switch willingly to a weaker assistant just because it is bundled into a managed iPhone. According to McKinsey’s 2025 AI in the Workplace report, 46 percent of business leaders already cite AI skill gaps as one of the biggest barriers to enterprise adoption. An assistant employees do not trust does not solve that problem. It makes it worse.
The Multi-Model Strategy Looks Smart – but Also Messy
Apple’s response to its AI gap increasingly looks like a platform move rather than a pure-model breakthrough. On the earnings call, Tim Cook confirmed that Google is set to play a major role in Apple’s AI strategy:
“Google’s AI technology would provide the most capable foundation for Apple Foundation Models.”
Reports suggest Apple is thinking beyond a single partner. The likely end state looks something like this: Siri powered at its core by Gemini, with users potentially able to tap ChatGPT, Claude, Perplexity, and other models as options on top.
MG Siegler argues that this may actually be the right consumer move:
“If they make it more of a first-class citizen – along with Claude if you wanted Claude, if you wanted Perplexity – they’ll just allow them to plug right into it on top of their baseline Siri powered by Gemini. I think that’s the right approach.”
For end users, that sounds flexible. For enterprise IT and security teams, it sounds more complicated. Which model is handling which query? What happens to data routing and residency when one employee uses Gemini and another uses OpenAI or Anthropic? Can an MDM policy standardise or restrict model choice across a managed fleet? Can those choices be audited?
These are not edge-case questions. According to PwC’s 2025 AI Agent Survey, 88 percent of senior executives plan to increase AI-related budgets in the next 12 months, yet data security and governance remain among the biggest blockers to enterprise-scale adoption. Apple has not yet published clear enterprise guidance on how a multi-model Siri would work in managed environments. Given the scale of corporate Apple deployments already underway, that silence stands out.
WWDC 2026 Is Now Apple’s AI Reckoning
Everything about Apple’s current AI narrative points toward June. WWDC 2026 feels unusually important because the company no longer has the luxury of talking about future potential alone.
Apple is certainly investing. R&D spending reached $18.4 billion in Q1 FY2026, up 19 percent year on year, driven in part by AI infrastructure, M-series silicon, and Private Cloud Compute. The hardware base is not the issue. M5 silicon in the latest iPad Pro and MacBook Air is already among the strongest foundations in the market for on-device AI workloads.
The problem is that the market has moved on. In June 2024, shipping a polished, context-aware AI assistant would have felt ambitious. In June 2026, the conversation is already about autonomous agents, workflow execution, and AI that can act on behalf of users rather than simply answer them.
That raises the bar dramatically. Apple does not just need to show a prettier Siri. It needs to show that its AI roadmap has a credible path into enterprise productivity, workflow automation, and governed execution.
Siegler’s concern is that Apple may still be short on the organisational muscle required to compete at that pace:
“There’s a real risk that they just don’t have the right mindset – and they are constantly running a step behind. And this is technology moving faster than any other that’s come before it.
What Enterprise IT Leaders Should Do Now
None of this makes Apple hardware a bad enterprise bet. The deployments at AstraZeneca, Snowflake, and Copel are not irrational. Apple’s device management story remains one of the strongest in the market, and M-series silicon is genuinely excellent. But the AI strategy needs to be evaluated separately.
- Decouple your device strategy from your assistant strategy. Apple hardware and Apple AI do not need to be the same decision. Running ChatGPT Enterprise, Microsoft Copilot, Gemini, or Claude on Apple devices is already common practice and entirely manageable in modern MDM environments.
- Watch WWDC for two concrete signals. First, does the new Siri experience ship with persistent history, admin controls, and governance detail that makes it enterprise-credible? Second, does Apple announce anything that resembles agentic capability, or is the focus still on conversational assistance alone?
- Push your Apple account team for specifics now. Ask how Apple Intelligence will work on enrolled devices, which models will be available through Siri, how data routing will be controlled, and what audit or logging capabilities will exist. These are not consumer questions. They are procurement questions.
- Do not assume hardware leadership automatically becomes AI leadership. The companies best positioned for the next productivity wave are building AI-native workflows now, regardless of which device sits at the endpoint. Apple’s strength is still the device. Right now, the intelligence layer is being built elsewhere.
Apple is not out of the race. The investment is real. The hardware foundation is strong. WWDC could still reset the conversation. But the clock is ticking, and enterprise buyers making multi-year endpoint decisions deserve more clarity on whether Apple’s AI ambitions will land before the market moves on without them.
To get the latest productivity and automation news, covering all the big enterprises, why not subscribe to our newsletter?
FAQs: Enterprise Agentic AI and Apple Intelligence
Is Apple Intelligence ready for enterprise use in 2026?
Apple Intelligence is getting closer, but it still does not look fully enterprise-ready on its own. There are useful features there already, including Writing Tools, Visual Intelligence, and Live Translation, and Apple hardware is clearly winning plenty of enterprise deployments. The bigger issue is Siri. It still lacks the persistent memory, governance controls, and agentic workflow capability that enterprise buyers are now starting to expect. That is why many IT leaders are still taking a split approach: Apple devices at the endpoint, but Microsoft Copilot, ChatGPT Enterprise, or Claude for Work as the real AI layer.
What is the difference between agentic AI and a regular AI assistant?
A regular AI assistant helps when you ask it to. It can answer questions, summarise a document, draft a reply, or pull together notes from a meeting. Agentic AI goes further. It can take action across multiple steps, systems, and workflows with much less hand-holding. Instead of only suggesting what should happen next, it can help make it happen. That is the real shift now underway in enterprise AI. The market is moving from assistance to execution, and that is why agentic systems from vendors such as OpenAI and Anthropic are getting so much attention.
Why is Apple behind in agentic AI compared to OpenAI and Anthropic?
The simplest answer is that Apple is still playing catch-up while others are already building the next layer. Rather than competing head-on in foundation models at the same pace as OpenAI or Anthropic, Apple has leaned more heavily on partnerships, especially around Google Gemini. That may prove sensible in some ways, but it also means Apple is not shaping this phase of AI from the front. While rivals are building systems designed for autonomous workflow execution, Apple is still trying to make Siri feel more credible as an assistant. In a market moving this quickly, that gap matters.
What should IT leaders look for from Apple at WWDC 2026?
There are really two things to watch. First, does Apple finally deliver a more capable Siri experience with features that feel serious enough for enterprise use, such as persistent history, better context handling, admin controls, and clearer governance detail? Second, does Apple show anything that looks genuinely agentic rather than just more polished assistant behaviour? That is the bigger test. WWDC 2026 is not just another product event. It feels more like a credibility moment for Apple’s enterprise AI roadmap.
Can enterprises use ChatGPT or Claude on Apple devices?
Yes, and plenty already do. ChatGPT Enterprise, Claude for Work, and Microsoft Copilot all work perfectly well on iPhone, iPad, and Mac, and can sit inside a standard managed-device environment. In practice, that is how many organisations are handling the gap today. They are buying into Apple hardware, while relying on third-party AI platforms to do the heavier lifting. Apple may eventually make that model more seamless through a multi-model Siri approach, but until the governance and control picture is clearer, most enterprise buyers will still want to treat that cautiously.
