Ollie bets on privacy protection to enter the competition with AI assistants
TechCrunch
1h ago
Ai Focus
AI Assistant Ollie Competes in the home scenario market by highlighting privacy and SOC compliance as selling points.
Helpful
No.Help

The competition among AI assistants aimed at family scenarios is intensifying. According to a report by TechCrunch, Ollie, headquartered in San Diego, is focusing on "privacy protection" as its main selling point, hoping to distinguish itself from a group of AI assistants that primarily assist with text messages and daily tasks.

Ollie currently focuses on family scheduling and lifestyle management. The product can connect to calendars and email to help users organize their family schedules, synchronize daily tasks, and also assist in planning meals, grocery shopping, tracking to-do items, booking services, and paying bills within group chats. The company also mentioned that future updates may include the addition of family budget management features.

Obtained SOC 2 compliance certifications

Ollie indicates that the company has met the SOC compliance standards. This is a framework that verifies through independent audits whether a company has established formal data protection and system security controls. For a AI assistant that needs to access users' schedules, emails, and payment processes, such certification is mainly used to prove to users that its data processing procedures have basic security guarantees.

Co-founder and CEO Bill Lennon stated that one of the purposes of adopting a subscription model for Ollie is to make it clear to users that the product and service are intended for paid users themselves, and not to use user data for other business purposes. According to him, the company will not share user data with third parties, nor will it use this data for other purposes, such as training models.

Distinguish with privacy and competitors

Currently, there are multiple participants in the personal AI assistant category, including text-based assistants such as Poke, Fambot, Ohai, Folk, Saner.ai, and Tomo. There are also products like Town, Lindy, and Reclaim.ai that focus more on tasks, calendar management, and email handling. Recently, Instinct, which has not yet been officially launched, has secured $350 million in financing and is valued at $2.5 billion, indicating that the popularity of this field is rising rapidly.

In contrast, Ollie has a smaller financing scale. Reports indicate that the company has completed a $7.5 million seed round of financing, with Khosla Ventures as the main investor, and AI House also participating in it.

Privacy issues are becoming an important area in this field. Reports mention that Instinct has previously been criticized for having overly broad service terms and privacy policies, which granted the platform extensive rights to access, store, transmit, publish, and modify users' data, and this data could also be used to train AI models. Ollie, on the other hand, is trying to highlight its differences by emphasizing that it "does not share data."

Login and payment still require user participation.

In terms of specific design, Ollie indicates that no username or password will be requested to complete tasks on behalf of the user. If logging into a website is necessary, the system will use its own browser in the cloud and send the user a remote session link, which the user will then log in to on their own. The process is roughly the same when it comes to payments or shopping.

This approach enhances security, but the trade-off is that users still need to participate repeatedly in some tasks. Lennon indicates that in the future, the company may explore safer tokenization solutions to reduce the steps of re-entering information, but at this stage, they are still seeking a balance between convenience and trust.

The report also pointed out that the stability of AI's assistants remains a common issue. TechCrunch mentioned during testing that Instinct once provided incorrect hotel prices, and Ollie also stopped responding for a time due to infrastructure failures with the SMS service provider. Lennon believes that large language models inherently possess probability and instability, and the product team needs to implement additional system designs to intercept and correct errors.

For Ollie, privacy and security can be reasons to attract users to give it a try, but when it comes to more sensitive scenarios such as accessing bank accounts and managing budgets, the reliability of the product will still determine whether users are willing to continue authorizing its use.

Tip
$0
Like
0
Save
0
Views 20
HKWDB reminds readers to view blockchain rationally, stay aware of risks, and beware of virtual token issuance and speculation. All content on this site represents market information or related viewpoints only and does not constitute any form of investment advice. If you find sensitive content, please click“Report”,and we will handle it promptly。
Submit
Comment 0
Hot
Latest
No comments yet. Be the first!
Related
web3: Ultrahuman Raises $70 Million in Financing to Advance AI Smart Rings
Ultrahuman Completes $70 Million in Financing and Collaborates with Qualcomm to Develop a New Generation of Smart Rings That Support Interaction with AI Local Software.
TechCrunch
·2026-09-04 01:23:37
10
web3 : Circular releases the Ring series, with new NFC payment and vibration alerts
Circular releases the Ring series of smart rings, featuring NFC payment, vibration alerts, and health tracking functions, with expected sale to begin in early 2027.
TechCrunch
·2026-09-04 00:58:32
22
web3: Sony反驳PS5 buyers' claim for a $508 million refund
Sony requests to dismiss the lawsuit for a refund due to the price increase of PS5, claiming that there is a lack of evidence for a causal relationship between tariffs and the price increase.
Coinpaper
·2026-09-04 00:45:30
23
Claude Fable 5.1 Turning Long-Term Tasks into the Main Battlefield: Prices Remain Unchanged, but Cache Costs Cut by 75% First
Anthropic launched Claude Fable 5.1 and Claude Mythos 5.1 on September 1st. Both use the same underlying model, with the main differences lying in access scope and security policies: Fable 5.1 is aimed at general professional users and developers, but it will implement restrictions or switch to a less capable model for high-risk requests in areas such as network security, biology, and chemistry; Mythos 5.1 retains more complete relevant capabilities and is only open to a small number of audited institutions.
CoinMeta
·2026-09-04 00:09:28
26
View More