• Google’s Pixel 11 series delivers a granular privacy and security experience, which closely aligns with survey findings around critical user requirements in regard to AI, agentic workflows and the relationship to personal, sensitive data. 
• Apple banks on its established history of privacy as a key differentiator for its user base. However, the new Siri app, and the integration of third-party AI services, illustrate Apple’s past focus on usage simplicity, instead of addressing clear consumer calls to have transparency and fine levels of control in the user experience. 
• For Samsung, the key challenge is the multitude of agentic AI services found on its flagship device. While some, like Galaxy AI, can be easily disabled, there is a lack of consistency bridging Gemini, Galaxy AI, or Bixby. 
• Motorola and OPPO do not have access to the most advanced automation features yet but should get many of them when they upgrade to Android 17. However, the current applications (Moto AI, OPPO Mind Space, and third-party services) create an environment where the lines between each service can become blurred, leaving the end-user in the dark.   

Agentic AI is changing the smartphone from a device that responds to commands into a platform that can anticipate needs, understand context and complete multi-step tasks across apps. That shift creates a major opportunity for smartphone OEMs, but it also raises the stakes for privacy, security, transparency and user control. As AI agents gain access to messages, calendars, photos, location, payments and work data, consumers will judge devices not only by what AI can do, but by how safely it does it. 

Counterpoint’s survey of 12,225 smartphone users across Australia, Germany, Japan, Mexico, the UK and the US found that consumers want AI to provide more personalized responses and are willing to let agents interact with personal data when strong safeguards are in place. But respondents also consistently pointed to privacy, transparency, control and security as non-negotiable requirements. The value of agentic AI depends on an exchange: users will share context if brands can prove that data is protected, visible and governed by the user. 

Priorities of AI and Personal Data

In connection with this in-depth survey, Counterpoint developed its 2026 Agentic AI Privacy & Security scorecard for mobile devices. 

Apple is extending its privacy-first brand promise into agentic AI through Siri AI, App Intents, View Annotations, Apple Foundation Models and Private Cloud Compute (PCC). The iPhone 17 series benefits from Apple’s established security model, local processing and tightly controlled hardware-software stack. However, the report also shows that Apple’s focus on simplicity can obscure what is happening in the background. Apple remains strong on security, but it must make AI activity easier for users to observe and control. 

Google is pushing agentic AI deeper into Android through Gemini, cross-app automation and real-time visibility tools, like progress-tracking interfaces. Its strongest advantage is the degree to which the Pixel experience can be controlled end to end, from on-device processing to cloud execution. This helps explain why the Pixel 11 series performs well across the scorecard. 

Samsung occupies a distinct position because its flagship experience combines Gemini, Galaxy AI, Bixby and partner-backed intelligence services. The Galaxy S26 Ultra benefits from Knox, KEEP, and Android Enterprise. It also gives privacy-conscious users a powerful “Process Data Only on Device” toggle for Galaxy AI. The challenge is consistency. Controls that apply to Galaxy AI do not necessarily apply to Gemini or Bixby, and the coexistence of multiple assistants can make it harder for users to know which service is acting, where data is going and what level of protection applies. 

Motorola and OPPO demonstrate the promise and risk of more open, partner-led AI approaches. Moto AI and OPPO Mind Space offer useful personalization and model-routing experiences, but they rely more heavily on third-party services and fragmented cloud and app ecosystems. Both present unique AI features and it will be critical for them to further develop and enhance them to create differentiating services to those provided by Google and Samsung. 

In the next phase of mobile AI, companies that turn privacy into a visible product feature will have a clear advantage over the competition. Security alone is not enough if users cannot see when an agent is active, understand which data it accesses, pause an automation or approve sensitive actions. Consumers are asking for AI that is personal, powerful and safe. The OEMs that can combine all three will be best positioned to make agentic AI a trusted part of everyday smartphone use. 

To evaluate how smartphone OEMs currently address key privacy and security features in their latest flagship devices, Counterpoint developed a scorecard evaluating three critical areas of concern for users. Data isolation and security, transparency and control emerged as key themes in a consumer study conducted by Counterpoint earlier this year.  

The survey surfaced that users believe: AI is integrated across most of the apps and features they currently use already; privacy and security are fundamental concerns about AI; virtual assistants are an every-day interaction tool. 

To evaluate how smartphone OEMs currently address key privacy and security features in their latest flagship devices, Counterpoint developed a scorecard evaluating three critical areas of concern for users. Data isolation and security, transparency and control emerged as key themes in a consumer study conducted by Counterpoint earlier this year.  

The survey surfaced that users believe: AI is integrated across most of the apps and features they currently use already; privacy and security are fundamental concerns about AI; virtual assistants are an every-day interaction tool.