A Wired reporter tested six AI assistants on real tasks for a day. They work better than last year, but connecting your email or bank carries real privacy risk.
A Wired reporter spent a day testing six AI assistants that act on your behalf, and found they follow through on tasks more reliably than last year. They still make mistakes. The tools are called agents, and the test aimed to show which ones are worth a nontechnical person's time.
The six were Meta's Muse, OpenAI's Dots, Instinct, Anthropic's Claude Duet, Google's Gemini Spark and xAI's Grok Bot. An agent is like a digital assistant you text. It can scan email, track flight deals or help with shopping, and it uses its own virtual browser to get online.
Each agent was scored from 1 to 5 across five tests. In the morning news briefing test, the reporter asked for 30 minutes of deep research. Only Muse and Dots kept researching for most of that time. Gemini Spark had the best story idea and earned 5 out of 5, though the editors who heard the pitches were not impressed overall.
The material available does not include the final scores or the overall winner. On privacy, agents work best when connected to email and calendars, which takes a lot of trust. One X user claimed his agent posted his bank balance where coworkers could see it. That is his claim, not a confirmed case. The reporter shared only Google Calendar and Gmail.
Start small with an AI agent: share only what you must, such as email or a calendar, and keep bank accounts and work chats out of it.
Lawmakers in both parties want Nvidia's CEO to answer questions about a reported tie to an alleged Chinese intelligence asset, and chip rules could shift as a result.
Read more: Lawmakers from both parties want Nvidia's CEO to answer questionsDeepSeek says 1.54 million people use its AI agent platform each day. If you run a business, cheaper AI helpers may soon be offered to you.
Read more: DeepSeek says its AI helper platform now has 1.54 million daily usersOpenAI has about $11 billion in hand at a $1.25 trillion value, below its first ask. Microsoft and NVIDIA shareholders have reason to watch what comes next.
Read more: OpenAI raises $11 billion, but at a lower price than it wanted
