Liability is not the only reason Google might choose to censor Covid keywords. They might do it because their users expect them to. They might do it because they think it's the right thing to do. They might do it because they care more about blocking disinformation than allowing useful information. They might do it because a programmer has been working on this really cool censoring algorithm and they want to test it in production. They might do it because they're tired of people talking about Covid on their platforms and want it shut down in general.

Section 230 applies to every major/minor platform in the US, granting all of them immunity from liability for most content. And yet, virtually all of them from Facebook to 4Chan, from Youtube to Pornhub, from Reddit to tiny 200-member Wordpress forums -- all of them censor some content. So liability is clearly not the only motivation for platforms to moderate, otherwise every comment section would be 8Chan.