A DESCRIPTION OF THE PROBLEM :
When using Files.readString(...) with UTF-16, encoding errors are ignored/ replaced, rather than throwing a MalformedInputException. This is the result of a fix of JDK-8286287, but is not matching the previous behavior of JDK 11 to JDK 16 and also inconsistent to the behavior for most other charsets, e.g. UTF-8.

REGRESSION : Last worked in version 16

STEPS TO FOLLOW TO REPRODUCE THE PROBLEM :
Run the test code.

EXPECTED VERSUS ACTUAL BEHAVIOR :
EXPECTED -
It should throw a MalformedInputException, as it does from JDK 11 to JDK 16. Note that when you change UTF_16 to UTF_8, the code does exhibit the expected behavior in all versions from JDK 11 to JDK 19.
ACTUAL -
It does not throw an exception with UTF_16 but silently proceeds with a replacement character (U+FFFD)

---------- BEGIN SOURCE ----------
import java.io.*;
import java.nio.file.*;
import java.nio.charset.*;

class Test {
  public static void main(String... arg) throws IOException {
    Path p = Files.write(Files.createTempFile("charset", "test"), new byte[] { -50 });
    try(Closeable c = () -> Files.delete(p)) {
        String s = Files.readString(p, StandardCharsets.UTF_16);
    }
  }
}

---------- END SOURCE ----------