May 28 (Reuters) - Motorsport's world governing body, the Paris-based FIA, is ramping up its use of artificial intelligence (AI) as it works with a global coalition of sports and governments to combat online abuse.

The first United Against Online Abuse (UAOA) conference is being held in the French capital on Wednesday to discuss measures to be taken.

The FIA has already partnered with Arwen AI, who have worked with the Mercedes, Red Bull and Alpine Formula One teams, and a spokesman said the F1 governing body's AI capability was being built up.

FIA president Mohammed Ben Sulayem told Reuters more than 4,000 sites had been blocked in the space of seven weeks last year before the FIA's annual congress.

The FIA spokesman said there had been discussions with TikTok, Google France and Facebook, Instagram and WhatsApp parent company Meta.

Attendees on Wednesday include representatives of World Rugby, the International Tennis Federation and ATP Tour, soccer body FIFA and France's minister for sport and the Olympics Amelie Oudea-Castera.

Ben Sulayem said a recent barometer survey revealed 75% of athletes had reported threats against their physical safety.

More than 90% agreed athletes would leave their sport if online abuse was not addressed. The problem also affects those who give their time for free.

"The volunteers. You remove them and there is no Formula One. No WRC (world rally championship). Nothing," said Ben Sulayem.

"I got threats, our volunteers got threats, our drivers got threats, our stewards. If we allow this to happen for long our sport will be beyond repair.

"To be united together at a government level or a media level or on federation level, I believe we can make a change," he added.

"We have the French government's support. To have the forces together I believe we can make a difference. I feel good about this, the momentum is there."