Apple’s ChatGPT integration raises data privacy and competition questions

The new Apple and OpenAI partnership aiming to ingrate ChatGPT into Apple’s operating systems, virtual assistant, and writing tools poses important questions about competition and privacy, experts told Euractiv.

Similar partnerships are already under the microscope and could face formal antitrust investigations.

Apple’s integration of ChatGPT could add to regulatory scrutiny about competition, including the risk of collusion among competitors sharing partnerships with the same Artificial Intelligence (AI) firm.

There are also worries about the impact on third-party app access and user privacy and processing shifts from on-device to cloud-based models, prompting scrutiny under the EU’s Digital Markets Act and data protection laws.

Euractiv contacted the UK’s Competition and Markets Authority (CMA), the US Federal Trade Council (FTC), the French competition authority (Autorité de la Concurrence), and the German Federal Cartel Office (Bundeskartellamt), but they did not reply or declined to comment.

Euractiv also reached out to Apple and OpenAI but they did not provide additional statements beyond those already available.

The integration

ChatGPT will be integrated into Apple’s operating systems for iPhones, iPads, and Mac computers later this year, according to the announcement, first made at the Worldwide Developer Conference 2024 on 8 June when introducing Apple’s newest operating system iOS18.

Users will “be able to use ChatGPT for free, without creating an account,” said Craig Federsghini, senior vice president of software engineering at Apple during the conference. The integration relates to OpenAI’s newest model, GPT-4o, which currently requires payment for access.

Siri, Apple’s virtual assistant, will be able to access ChatGPT, as well as user messages and emails, to provide more accurate and contextual responses, the company said. Generative AI features integrated with Apple writing tools like Notes and Pages will help users create written and visual content.

Competition concerns

Regulators could scrutinise potential anticompetitive behaviour through the lens of antitrust laws or the EU’s Digital Markets Act (DMA).

OpenAI is already under the microscope in the EU, UK, and US for its partnership with Microsoft. Other similar partnerships, such as Amazon’s with Anthropic, are in a similar position.

These deals may not fall under existing merger rules, but could still face scrutiny over competition concerns. For example, the EU antitrust regulator will reportedly not investigate the OpenAI-Microsoft partnership under merger rules but may still investigate Microsoft for potentially distorting competition within the internal market, focusing on whether its market power leads to anti-competitive practices.

Some policymakers worry that big tech firms will dominate AI, through these acquisitions and partnerships with AI startups, despite not being subject to merger rules, said Lazar Radic, senior scholar for competition policy at the International Center for Law & Economics, and adjunct professor of law at IE University.

Under the DMA, Apple’s designated “gatekeeper” status requires it to ensure fair access and prevent practices that could stifle competition, including those that disadvantage startups.

The DMA and its processes are “quite flexible” and “can be updated if needed,” Commission spokesperson Lea Zubler said at a Tuesday (11 June) press conference, when asked by Euractiv about potential scrutiny of the Apple-OpenAI integration.

A future concern could also be collusion among competitors who share partnerships with the same AI company, such as Apple and Microsoft with OpenAI, said Radic.

“It is important to ensure that these companies do not exchange sensitive information that could distort competition,” he said.

There are also questions about how Apple will make sure other apps can work with its AI tools, especially those that compete with Apple’s own, and how its AI using data from other apps will affect how users interact with those apps, which is important for apps that rely on ads, an antitrust lawyer told Euractiv, on condition of anonymity.

Data privacy

Apple said the integration will not compromise the privacy of Apple OS users in its statement. OpenAI doesn’t store user requests and “users’ IP addresses are obscured”, said the firm.

But the use of such techniques, like masking IP addresses, can be a “hazy area,” Arnav Joshi, a senior technology lawyer also at Clifford Chance, told Euractiv. Another hazy area is “whether true, bar-raising and privacy-preserving options are really viable at scale – practically and legally, I think we’re in uncharted territory there,” he said.

Daniel Leufer, senior policy analyst at Access Now told Euractiv, that Apple has “shifted away from its relatively high privacy standards”, while Joshi disagreed: “Apple has a strong track record on user privacy, and bases a lot of its marketing on it, so it knows what’s at stake here.”

The firms emphasised consent in the announcements, saying “Apple users are asked before any questions are sent to ChatGPT, along with any documents or photos, and Siri then presents the answer directly.”

However, obtaining consent may not solve all issues when launching new features. Consent and opt-ins are complex under European data protection law (GDPR) and may not always be the best approach for introducing new features, Joshi said.

OpenAI’s compliance with the EU’s General Data Protection Regulation (GDPR) has been patchy. The Italian Data Protection Authority reopened an investigation into the company in March and the European privacy regulators launched a dedicated task force to address privacy in ChatGPT last year.

[Edited by Eliza Gkritsi/Zoran Radosavljevic]

Read more with Euractiv