Summary: OpenAI Academy and Older Adults Technology Services from AARP ran the free in-person Older Adults AI Skills Jam on September 16, 2026. The program teaches older adults across ten US communities to use ChatGPT for everyday tasks while recognizing urgent language, secrecy, suspicious links, and other scam warning signs. OpenAI says participant questions and experiences will inform future learning resources, but attendance, curriculum length, and outcome measures were not provided.
The people who decide how an AI product works tend to be people for whom a new interface is self-evident. That assumption breaks the first time your user is 72 and has never once been asked to phrase a request as a prompt.
According to OpenAI’s announcement, OpenAI Academy and Older Adults Technology Services (OATS) from AARP ran the Older Adults AI Skills Jam on September 16, 2026 — a free, in-person workshop that teaches older adults to use ChatGPT with more confidence and safety. The Jam sits inside a multi-year effort delivered through OATS’s Senior Planet program.
What the sessions actually teach
The framing stays concrete, which is refreshing. Trip planning, understanding a confusing letter or bill, checking whether something is a scam, picking up a new hobby, staying in touch with family — that is the list the announcement gives.
Safety has a block of its own. Participants go through the standard warning signs — urgent language, secrecy, suspicious links — plus one rule stated plainly in the announcement: pause, think, and ask. ChatGPT is framed as an extra layer of scrutiny, not a replacement for judgment.
For anyone building consumer AI, that distinction carries weight. “Ask the model if this is a scam” is a very different product promise from “the model will protect you.” One keeps the user in the loop; the other sets an expectation no model reliably meets.
The numbers that explain the curriculum
Data from OpenAI itself, cited in the same post, shows the share of US messages tied to people 55 and older climbing from 6% to nearly 10% across a year. The announcement adds that people turn to ChatGPT for help evaluating suspicious messages, emails, and websites tens of millions of times a week.
Set those two figures side by side and the reason for the program becomes obvious. Older adults already use the tool for precisely the tasks the workshop teaches — the gap isn’t adoption, it’s fluency and risk awareness.
Ten communities, each with a local partner
Ten communities host the Jam, each one paired with an organization from the area: Senior Planet sites in Denver, Miami, San Antonio, Montgomery County, and Queens, along with the Mirowitz Center in St. Louis, Senior Community Services in the Twin Cities, Nashville Public Library with FiftyForward, Fresno EOC, and LEARN Idaho in Boise.
Similar events from OpenAI Academy have served nonprofits, small businesses, and K–12 educators. The older-adult track is the latest addition, and according to the announcement, the questions and experiences of participants will feed back into future learning resources.
What changes if older users are your audience
For builders, three things stand out.
First, onboarding is a product surface rather than documentation. A workshop that teaches “pause, think, and ask” is doing work your empty state could do too. If the first screen of your assistant assumes the user knows what a good prompt looks like, you’ve handed that lesson off to a library in Boise.
Second, scam-checking is a genuine use case, and it arrives with a genuine failure mode. When someone pastes a suspicious email, the reply has to be legible and hedged the right way — not delivered as a confident verdict. The trust patterns at work here look like the ones in Fyxer’s draft acceptance research: an assistant earns trust by being right at the moments a user can check, not by sounding certain everywhere.
Third, for this audience in-person beats self-serve. Ten physical locations staffed through trusted local partners add up to a distribution strategy, not a marketing stunt. If your growth model assumes users will work it out on their own, that assumption is worth testing against a cohort that didn’t grow up with chat interfaces.
What the supplied announcement leaves out: attendance numbers, curriculum length, and how outcomes will be measured. Those details would say more about whether the format works — and whether it’s worth copying.