Microsoft and Apple will not have advisory roles on OpenAI’s board of directors as regulatory scrutiny on Big Tech companies increases. Despite Microsoft’s significant investment in OpenAI, the company will not have a nonvoting seat on the board. Apple’s plan to take an advisory board role has also been canceled. OpenAI confirmed that its board will have no advisory seats in the future.
While that’s the headline, let me do a bit more OpenAI updates.
A hacker breached OpenAI’s internal messaging systems stealing details about the design of the company’s AI technologies in an incident discussesd at an-all hands meeting in April 2023. While no customer or partner information was stolen, concerns were raised about the potential for foreign adversaries, particularly China, to steal AI technology that could pose a threat to national security. OpenAI executives did not publicly disclose the breach but faced internal criticism regarding security measures.
OpenAI’s ChatGPT macOS app had a security issue where conversations were stored in plain text, making them easily accessible to bad actors or malicious apps. After being alerted to the issue, OpenAI released an update that encrypts the chats to address the problem. The app no longer allows access to conversations in plain text. OpenAI may review ChatGPT conversations for safety and training purposes, but this access should not extend to unknown third parties.
Why do we care?
Board governance is becoming a significant factor to the vendors that service providers are using, and worth keeping an eye on.
The breach of OpenAI’s internal messaging systems and the subsequent internal criticism raise serious concerns about cybersecurity in AI development. With the potential for foreign adversaries to exploit stolen AI technology, this highlights why CISA’s initatives to drive security into product development matter. Add a dash of lack of security in their new macOS app to that why.