Apple Tightens Mac Security as AI Agents Go Rogue, Meta Opens Muse to DIY Gadgets
Today's roundup sees Apple taking a firm stance on data privacy, tightening macOS security settings after concerns arose about AI agents accessing private messages. Meanwhile, Meta is inviting developers to build their own hardware for its AI assistant, Muse. We also look at Stability AI's new direction, focusing on music generation.
In this round-up
- Apple clamps down on Mac full disk access for AI agents
- Meta allows DIY hardware integration for its Muse AI agent
- OpenAI launches enterprise-focused agent platform called Dots
- Stability AI pivots to music generation under Sean Parker
- White House rebrands AI as 'super intelligence', tech leaders sign safety pledge
Apple clamps down on Mac full disk access for AI agents
Apple is updating its macOS privacy settings to stop third-party AI applications from misusing full-disk access permissions. This follows an incident where Meta's Muse AI agent reportedly accessed a user's private messages without explicit permission, sparking an uproar on social media. Apple stated that some developers have been using full-disk access in ways that expose user data, including files, mail, messages, and browsing history, without their full knowledge. The company emphasised that as AI agents become more capable and autonomous, the risks associated with this level of access will increase substantially. The new controls will require very explicit user action to grant such broad permissions.
For UK event professionals, this means increased scrutiny on any AI tools they use on Mac devices that request broad system permissions. Data privacy and security are paramount, especially when handling delegate information, supplier contracts, or sensitive event plans. Ensure that any AI software integrated into your workflow adheres to strict data protection standards and that you understand exactly what permissions you are granting. This also highlights the importance of keeping abreast of privacy settings on your devices and questioning broad access requests from any application.
Read more: theverge.com · techcrunch.com
Meta allows DIY hardware integration for its Muse AI agent
Meta has launched Muse Gadgets, an open-source project that allows developers and hobbyists to build their own hardware to connect with Muse, its personal AI agent. The company is providing open-source firmware and a Linux software development kit, along with project ideas like adding Muse to a colour e-ink display or an HDMI stick for TV integration. Meta itself built a USB-C powered device called Muse Home Link, enabling Muse to connect to home networks and smart devices. Five thousand of these Home Links are being given away to Muse subscribers. This move aims to make Muse more than just a chatbot, extending its capabilities into various physical devices and environments.
For event professionals, this could open up interesting possibilities for bespoke AI integrations at events. Imagine custom-built devices that offer on-site delegate support, interactive displays, or smart room controls powered by Muse. While this is primarily a developer-focused initiative for now, it signifies a trend towards more embedded and pervasive AI. There is nothing to do right now, but it is worth keeping an eye on how these hardware integrations evolve, as they could eventually lead to new forms of interactive experiences or efficient operational tools for events.
Read more: techcrunch.com · theverge.com
OpenAI launches enterprise-focused agent platform called Dots
OpenAI has introduced a new agent platform named Dots, which is presented as enterprise software capable of assisting with work tasks and personal errands. Dots feature anthropomorphic avatars and an interface similar to Meta's Muse, where users interact with the agent in one window and observe its actions in a virtual machine in another. These agents can access applications like Blender and GIMP and can also be connected to a user's computer via the desktop ChatGPT app. OpenAI is initially rolling out Dots to its highest-tier subscribers, including the $100 per month Pro account, indicating a business-centric approach.
For event professionals, Dots represent another step towards AI agents becoming integral to business operations. Unlike Meta's consumer-friendly Muse, OpenAI's Dots appear positioned for more complex, work-related automation. This could mean using an AI agent to manage aspects of event planning software, administrative tasks, or even complex design work. While the high cost and enterprise focus might mean it is not for everyone yet, it is a clear signal of the direction AI is heading for business users. Keep an eye on its capabilities, as this type of agent could eventually streamline many back-office functions for event agencies and venues.
Read more: theverge.com
Stability AI pivots to music generation under Sean Parker
Stability AI, known for its image generation technology, is being rebuilt by Napster co-founder Sean Parker with a new focus on music. Parker, who joined a $80 million rescue of the company two years ago, along with CEO Prem Akkaraju, is steering Stability AI to become a key AI toolmaker for music professionals. The company secured $76 million in funding from entities including Sony, Warner, and Universal, which also licensed their music catalogues for training purposes. Stability AI has already released three new audio models and AI music-editing software, capable of generating instrumental tracks or short snippets from text prompts. An upcoming update will allow users to hum melodies or beatbox to guide the AI.
For the events industry, this development could have significant implications for event production and entertainment. Event organisers and production companies might soon have access to sophisticated AI tools that can generate custom music for specific event themes, background tracks, or even interactive soundscapes. This could reduce reliance on stock music and offer new creative possibilities for bespoke audio experiences. While still emerging, it is worth watching how this technology develops and if it becomes a cost-effective solution for unique event sound design.
Read more: techcrunch.com
White House rebrands AI as 'super intelligence', tech leaders sign safety pledge
The White House has officially rebranded AI as 'super intelligence' following an executive order signed by President Donald Trump. This comes after nearly every major tech CEO, including Mark Zuckerberg, Jeff Bezos, Elon Musk, and Anthropic's Dario Amodei, gathered at the White House to sign an AI safety pledge that President Trump described as 'morally binding'. The move highlights increasing government attention on the development and deployment of advanced AI, with a focus on safety and ethical considerations.
For event professionals, this renaming of AI to 'super intelligence' is largely a branding exercise from the US government, but it signifies the growing public and political importance of AI. While it does not change the practical application of AI tools in events, it underscores the need for businesses to be aware of the broader societal conversations around AI's capabilities and risks. It is a reminder that as AI becomes more integrated into daily life and business, its ethical use and safety will remain a key concern, which could influence future regulations and public perception of AI-powered event technologies.
Read more: techcrunch.com · techcrunch.com
Sources
- Apple will limit Mac disk access as AI agents ‘substantially’ increase risk
- Apple says it’s tightening macOS ‘Full Disk Access’ controls due to new risks from AI agents
- Meta wants your next gadget to be Muse-infused
- Meta open sources code to let you make Muse AI gadgets
- OpenAI’s Dot agent is enterprise software that can also order your dinner
- Sean Parker is rebuilding Stability AI around music
- Call it AI, call it Super Intelligence, only 2% of consumers are buying it
- It’s not AI anymore, it’s ‘super intelligence’ (according to the White House)