[Bug] Geom_smooth graphical error when converting from ggplot to plotly using ggplotly

Hello,

I have a graphical issue which shows up when converting a ggplot2 plot with geom_smooth to a plotly plot:

as you can see in the plotly plot, an extra nonsensical confidence interval ribbon gets generated below the fitted lines.

When I lower the confidence interval from 0.95 to 0.80 the ribbon in correctly displayed:

However this is not the confidence interval I need.
I know it has something to do with plotly as this is my ggplot (with confidence interval 0.95) used to generate the first bugged image:

Here is the (anonymized) code used to generate the plot(s):

p16 <- ggplot(data= df, order = rownames(x)) +
geom_point(data= df,aes(x=x.ordered, y=(mean1), color=paste("Number = ", Number))) +
geom_point(data= df,aes(x=x.ordered, y=(mean2), color=paste("Number2 = ", Number))) +
geom_smooth(data = dfcopy, aes(x=x.ordered, y=mean2, color=Number, group=Number), method="loess", level=0.95) +
ylab("y") + scale_y_continuous(limits=c(0, 1)) + xlab(paste("x")) +
theme_bw() + theme(plot.title = element_text(hjust = -.5, face="bold"), axis.title.x = element_text(hjust = 1, size = 14, face="bold"), axis.text.x = element_text(angle = 90, hjust = 1), axis.title.y = element_text(size = 14, face="bold")) +
ggtitle("") + labs(color = "\n\nType of data point: " )

p16 <- p16 + scale_color_manual(values=c("#B24745FF", "#DF8F44FF","#00A1D5FF", "#6A6599FF", "#00A1D5FF", "#6A6599FF"))
gp16 <- ggplotly(p16, tooltip=c("text")) %>% layout(legend = list(orientation = "v",xanchor = "center", x = 1.2), title = list())

Unfortunately I can't provide data or the axes, I also removed the ggplot text variable due to anonymous data.

If someone might know a solution to this graphical error, it would be greatly appreciated.

No repository file or test is named, and the reporter cannot provide the data or axes. Start at the ggplotly conversion path with a reproducible geom_smooth loess plot using level=0.95, comparing it with the ggplot output. Done means the extra confidence-interval ribbon is no longer generated and the converted plot matches the source plot.