Zero protection from spoofing, which I think is a worry.
What we see in very significant volumes at the moment (globally, not just UK) is threat-actors using fairly basic credential phishing attacks/brute force attacks to compromise ill-defended user accounts.
An example of how this would work here would be that the threat-actor successfully compromises a user account at the AMG Brabus Collection, our research shows that within 30 seconds the account will have been accessed and searched for any financial interaction - search terms being run such as 'account', 'payment', 'invoice' and so forth are the very first actions we see in most compromised accounts.
In this case if ABC are selling cars the threat-actor would find conversations where the deposit has been paid but not the balance.
Next action the threat-actor takes is to send an email to the person buying the car spoofing the actual domain, with either a convincing look-alike (or more commonly for low value attacks) just a throw-away address in the reply-to.
This email would have the previous exchanges in the conversation pasted into it, would be put into the thread as a genuine reply, and would likely pass inspection - and would ask the buyer to send the balance to a different account (if details already provided) or would ask for the balance via bank transfer to a specific account (if not).
Buyer transfers the balance, a few days later they turn up at ABC to collect their car - but of course they've not paid for it.