Equivalent of utf8_general_ci in Postgres/ICU?

In MySQL there is a collation utf8_general_ci which provides case-insensitive comparisons in a variety of languages. For example, these are all 1 (true):

SELECT 'ı' = 'I' SOLLATE 'utf8_general_ci';
SELECT 'i' = 'I' COLLATE 'utf8_general_ci';
SELECT 'ä' = 'Ä' COLLATE 'utf8_general_ci';

Can I define a similar collation using PostgreSQL's ICU? I tried

CREATE COLLATION "undefined_ci_nondet_old" (
  PROVIDER = 'icu',
  LOCALE = "@colStrength=secondary",
  DETERMINISTIC = false
);

But that doesn't seem to include the Turkish I/ı conversion:

SELECT 'ı' = 'I' COLLATE undefined_ci_nondet_old; -- false