refactor(fuzz): remove custom fuzz_utils module

diff --git a/fuzz/fuzz_targets/bitcoin/deserialize_psbt.rs b/fuzz/fuzz_targets/bitcoin/deserialize_psbt.rs
index 7b36cf69..c0d73691 100644
--- a/fuzz/fuzz_targets/bitcoin/deserialize_psbt.rs
+++ b/fuzz/fuzz_targets/bitcoin/deserialize_psbt.rs
@@ -1,29 +1,30 @@
-use bitcoin_fuzz::fuzz_utils::consume_random_bytes;
+use arbitrary::{Arbitrary, Unstructured};
 use honggfuzz::fuzz;
 
 fn do_test(data: &[u8]) {
-    let mut new_data = data;
-    let bytes = consume_random_bytes(&mut new_data);
-    let psbt: Result<bitcoin::psbt::Psbt, _> = bitcoin::psbt::Psbt::deserialize(bytes);
-    match psbt {
-        Err(_) => {}
-        Ok(mut psbt) => {
-            let ser = bitcoin::psbt::Psbt::serialize(&psbt);
-            let deser = bitcoin::psbt::Psbt::deserialize(&ser).unwrap();
-            // Since the fuzz data could order psbt fields differently, we compare to our deser/ser instead of data
-            assert_eq!(ser, bitcoin::psbt::Psbt::serialize(&deser));
-
-            let new_bytes = consume_random_bytes(&mut new_data);
-            let psbt_b: Result<bitcoin::psbt::Psbt, _> =
-                bitcoin::psbt::Psbt::deserialize(new_bytes);
-            match psbt_b {
-                Err(_) => {}
-                Ok(mut psbt_b) => {
-                    assert_eq!(psbt_b.combine(psbt.clone()).is_ok(), psbt.combine(psbt_b).is_ok());
-                }
-            }
-        }
-    }
+    let mut unstructured = Unstructured::new(data);
+
+    let Ok(bytes_a) = <&[u8]>::arbitrary(&mut unstructured) else {
+        return;
+    };
+    let Ok(bytes_b) = <&[u8]>::arbitrary(&mut unstructured) else {
+        return;
+    };
+
+    let Ok(psbt_a) = bitcoin::psbt::Psbt::deserialize(bytes_a) else {
+        return;
+    };
+
+    let ser = bitcoin::psbt::Psbt::serialize(&psbt_a);
+    let deser = bitcoin::psbt::Psbt::deserialize(&ser).unwrap();
+    assert_eq!(ser, bitcoin::psbt::Psbt::serialize(&deser));
+
+    let Ok(mut psbt_b) = bitcoin::psbt::Psbt::deserialize(bytes_b) else {
+        return;
+    };
+
+    let mut psbt_a_clone = psbt_a.clone();
+    assert_eq!(psbt_b.combine(psbt_a).is_ok(), psbt_a_clone.combine(psbt_b).is_ok());
 }
 
 fn main() {
diff --git a/fuzz/src/fuzz_utils.rs b/fuzz/src/fuzz_utils.rs
deleted file mode 100644
index 883729c3..00000000
--- a/fuzz/src/fuzz_utils.rs
+++ /dev/null
@@ -1,50 +0,0 @@
-// SPDX-License-Identifier: CC0-1.0
-
-//! Helper functions for fuzzing.
-
-pub fn consume_random_bytes<'a>(data: &mut &'a [u8]) -> &'a [u8] {
-    if data.is_empty() {
-        return &[];
-    }
-
-    let length = (data[0] as usize) % (data.len() + 1);
-    let (bytes, rest) = data.split_at(length);
-    *data = rest;
-
-    bytes
-}
-
-#[allow(dead_code)]
-pub fn consume_u64(data: &mut &[u8]) -> u64 {
-    // We need at least 8 bytes to read a u64
-    if data.len() < 8 {
-        return 0;
-    }
-
-    let (u64_bytes, rest) = data.split_at(8);
-    *data = rest;
-
-    u64::from_le_bytes([
-        u64_bytes[0],
-        u64_bytes[1],
-        u64_bytes[2],
-        u64_bytes[3],
-        u64_bytes[4],
-        u64_bytes[5],
-        u64_bytes[6],
-        u64_bytes[7],
-    ])
-}
-
-#[allow(dead_code)]
-pub fn consume_u32(data: &mut &[u8]) -> u32 {
-    // We need at least 4 bytes to read a u32
-    if data.len() < 4 {
-        return 0;
-    }
-
-    let (u32_bytes, rest) = data.split_at(4);
-    *data = rest;
-
-    u32::from_le_bytes([u32_bytes[0], u32_bytes[1], u32_bytes[2], u32_bytes[3]])
-}
diff --git a/fuzz/src/lib.rs b/fuzz/src/lib.rs
deleted file mode 100644
index 3421bc22..00000000
--- a/fuzz/src/lib.rs
+++ /dev/null
@@ -1,5 +0,0 @@
-// SPDX-License-Identifier: CC0-1.0
-
-//! Fuzzing
-
-pub mod fuzz_utils;