To some degree, thorium is snake oil. They have some interesting properties and there are designs proposed that feature passive safety systems, but they remain largely unproven and they aren't the magic that people seem to often claim.
When it comes down to it, power generation is always risky. Reasonable efficiency demands that any power plant tries to maximize energy density, which is fundamentally in conflict with safety. Even solar and wind have risks, though obviously they are smaller than nuclear, coal or natural gas (though wind and solar are a different class of power generation and don't necessarily compete with those).
The main risks in nuclear power are loss of coolant, containment breach or a prompt criticality event. Thorium reactors are vulnerable to all of the above, but of course designs exploit various properties to mitigate the risk. The most famous thorium design is the liquid salt design, which is not as susceptible to loss of coolant flow accidents due to a reactivity feedback mechanism (it's subcritical if the liquid fuel stops being circulated), but it isn't inherently superior to other modern passively safe designs.
The real reason to be interested in thorium is that it has reduced proliferation concern and also thorium is available in places where uranium is rare (India). But it isn't fundamentally safer.