You can't pass data back to the original synchronous context once you have started an async context, as this would require the original context to "block" while it waits for the asynchronous results. Swift does not allow blocking in its concurrency model as this could lead to a thread deadlock.

You will have to just call another function with the results from the Task context to process the returned values. It's up to you if this process is another async function or not, depending on what you need to do.

func getReviewIDs() {
    Task {
        let result = await getReviewIdsFromGoogle()
        process(ids: result)
    }
}

func process(ids: [Int]) {
    print("now process ids: \(ids)")
}

func getReviewIdsFromGoogle() async -> [Int] {
    await withUnsafeContinuation { continuation in
        DispatchQueue.main.asyncAfter(deadline: .now() + 2.0) {
            continuation.resume(returning: [1,2,3])
        }
    }
}

To prevent data races you must use synchronized access to variables from concurrent operations and the compiler doesn't allow you to change your array directly. To avoid the issue you can implement isolated access to your data with an actor instance e.g.:

actor Store {
    var reviewIds: [Int] = []
    func append(ids: [Int]) {
        reviewIds.append(contentsOf: ids)
    }
}

func getReviewIds() {
    
    let store = Store()
    
    Task {
        let ids = await getReviewIdsFromGoogle()
        await store.append(ids: ids)
        print(await store.reviewIds)
    }
}