Just when we thought we were getting the hang of ChatGPT, OpenAI introduced something new. Meet Dot, an AI assistant that doesn't just answer questions but can help keep your work moving, even when you're busy doing something else.
Remember when ChatGPT first arrived and we were all amazed that it could write an email, explain a complicated spreadsheet or come up with a clever LinkedIn caption? Fast-forward a few years, and we're using AI to analyse data, build websites, automate workflows and create everything from marketing campaigns to business strategies.
Just as we're getting comfortable with all these capabilities, OpenAI has introduced another interesting development. What if, instead of constantly telling AI what to do next, we could give it a goal and let it get on with the job?
That's the thinking behind Dot, OpenAI's new proactive AI assistant. Think of it as a potentially very productive new colleague who doesn't need coffee, annual leave or a gentle reminder about that report you requested last Thursday.
And while that might sound a little futuristic, it's an interesting indication of where AI is heading.
Announced by OpenAI on 29 September 2026, Dot is an always-on AI agent designed to help people manage ongoing work rather than simply respond to individual requests.
Traditional ChatGPT is largely reactive. You ask it a question, give it instructions or request help with a task, and it responds. Dot takes this relationship further by remembering the context of your work, using connected applications and continuing to make progress between conversations.
Powered by GPT-6 Astra, Dot has its own cloud-based computer and can connect to thousands of applications through OpenAI's plugin ecosystem. It can research information, prepare documents, follow developments and bring completed work back for review. You decide which applications it can access, what responsibilities to give it and which actions require your approval.
In simple terms, it's the difference between having a very clever tool that waits for instructions and having an assistant that knows what you're working towards.
This is part of a broader shift towards AI agents, something we've explored in our article on how AI agents are changing the workplace and traditional organisational structures. As these technologies become more capable, businesses are beginning to rethink not only which tasks can be automated but also how people and AI can work together.
It's 08:00 on a Monday morning. You've barely opened your laptop, your coffee is still brewing, and your inbox already looks like it has a personal vendetta against you.
Normally, you'd spend the first hour figuring out what happened on Friday, which clients need attention, what deadlines are approaching and whether anyone followed up on those leads from last week. By the time you've worked through your inbox, checked your calendar and opened a few dashboards, a good chunk of your morning has disappeared.
Now imagine opening your laptop to find that your AI assistant has already made a start. Your weekly marketing report has been prepared for review, a handful of sales opportunities that need attention have been flagged, and a briefing for your upcoming client meeting is waiting. Dot has also noticed an unusual change in campaign performance and highlighted it so you can investigate before your next meeting.
These are examples of the types of workflows proactive AI could support, depending on the information and applications available to it. Dot won't automatically know everything about your business, but it can help reduce the amount of time spent gathering information and remembering what needs to happen next.
And frankly, who wouldn't appreciate a little help with Monday mornings?
For years, we've treated AI as something we open when we need help. Need a blog outline? Ask ChatGPT. Struggling with an Excel formula? Ask ChatGPT. Need to make a slightly passive-aggressive email sound professional? ChatGPT has probably saved a few workplace relationships by now.
The common thread is that we're usually responsible for initiating the interaction. We decide what needs doing, explain the task and return when we need something else.
Dot introduces a different way of working. Instead of treating AI as an occasional assistant, you can give it an ongoing goal and allow it to work towards that goal over time. It can retain relevant context, investigate developments and prepare useful work without needing a brand-new conversation for every step.
For example, rather than asking for a fresh campaign analysis every Friday, a marketing manager could give Dot an ongoing responsibility to help review performance. With the appropriate connections and permissions, it could monitor available information, identify unusual changes and prepare observations for the manager to consider.
The interesting development isn't simply that AI can complete more tasks. It's that we're moving towards technology that can help manage the work between those tasks, reducing some of the administrative effort that comes with keeping everything on track.
This evolution is closely connected to the broader movement from experimenting with individual AI tools towards building AI into everyday business operations. It's a topic we've also explored through OpenAI's introduction of AgentKit and the growing role of AI agents.
This is where things become particularly interesting for businesses.
Most marketing and sales teams aren't exactly short of technology. Between CRM platforms, analytics dashboards, project management tools, advertising accounts and email systems, there's certainly no shortage of places to log in.
The real challenge is often getting those systems to work together and finding enough time to make sense of all the information they produce.
Imagine a marketing team running campaigns across Google, Meta and LinkedIn while managing leads through HubSpot. Each platform provides valuable information, but someone still needs to analyse performance, identify opportunities and decide what happens next.
A proactive AI assistant could help bridge some of those gaps by reviewing connected data, identifying patterns and preparing recommendations for the team. Instead of spending hours pulling information together, marketers could spend more time interpreting the results and deciding where to focus their efforts.
The possibilities extend well beyond campaign reporting. For content teams, Dot could help turn an interview transcript into a blog outline, prepare supporting social posts and carry feedback across multiple drafts. Sales teams could use it to research prospects, prepare meeting notes and identify outstanding questions before a follow-up conversation. Operations teams could benefit from an assistant that keeps track of project developments and highlights tasks requiring attention.
These applications build on the same principles behind scaling sales outreach through RevOps, AI and connected CRM systems. When information flows between systems and repetitive processes are automated, teams have more time to focus on the activities that actually contribute to business growth.
Dot began rolling out on 29 September 2026, but it isn't available to every ChatGPT user.
OpenAI is introducing the feature gradually to eligible Pro users in supported markets and qualifying Business Premium workspaces. Enterprise organisations can also access a beta, subject to administrator approval and workspace requirements.
Users with access can create a Dot through ChatGPT, give it a name, connect supported applications and establish the work they'd like it to help with. From there, they can review its activity, provide feedback and adjust its responsibilities.
Availability, functionality and usage allowances may change as OpenAI continues developing the product. For the latest information, visit OpenAI's official Dot announcement or its getting-started guide.
For businesses that don't have access yet, there's still value in understanding what Dot represents. The technology offers a glimpse into how AI could become more closely integrated into everyday operations, rather than remaining a separate tool employees use occasionally.
There's something undeniably appealing about having an AI assistant that can work in the background, identify opportunities and take repetitive tasks off your plate. But before businesses rush to give their new digital colleague a name and a list of responsibilities, there's an important question worth asking: are the foundations ready?
AI can only work effectively with the information, systems and processes available to it. If your CRM is full of duplicate contacts, your reporting is inconsistent and your sales and marketing teams are working from completely different information, introducing a proactive AI assistant won't magically fix those issues.
In some cases, it could simply help you make the same mistakes faster.
At Velocity, we believe the real opportunity lies in combining AI with reliable data, connected systems and well-designed processes. Businesses that invest in these foundations will be better positioned to take advantage of increasingly capable AI tools, whether that's Dot or the next development around the corner.
This is also why the conversation around AI shouldn't revolve entirely around replacing people or automating every possible task. The more valuable question is how technology can free teams from repetitive administration and give them more time for the work that genuinely requires human expertise.
When AI handles more of the information gathering, monitoring and preparation, people can focus on strategy, creativity, customer relationships and decision-making. That's where the real business value starts to become interesting.
It's probably too early to say exactly how much Dot will change the way we work, but the direction is becoming increasingly clear. We're moving from AI that primarily responds to instructions towards AI that can take on ongoing responsibilities, understand context and help keep projects moving.
There will undoubtedly be some teething problems, and businesses will need to find the right balance between automation and human oversight. But the prospect of arriving at work to find that some of your most tedious tasks have already been taken care of is certainly appealing.
And if Dot can eventually figure out which meetings could have been emails, we might really be onto something.
AI is evolving quickly, but making it genuinely useful takes more than signing up for the latest platform. At Velocity, we help businesses connect their data, streamline processes and introduce practical AI and automation solutions that support real business goals.