diff --git a/ntcore/src/main/native/cpp/net3/WireDecoder3.cpp b/ntcore/src/main/native/cpp/net3/WireDecoder3.cpp
index 03bc9c11c9c..dea6349b010 100644
--- a/ntcore/src/main/native/cpp/net3/WireDecoder3.cpp
+++ b/ntcore/src/main/native/cpp/net3/WireDecoder3.cpp
@@ -5,11 +5,11 @@
 #include "WireDecoder3.h"
 
 #include <algorithm>
+#include <bit>
 #include <string>
 #include <utility>
 
 #include <fmt/format.h>
-#include <wpi/MathExtras.h>
 #include <wpi/SpanExtras.h>
 
 #include "Message3.h"
@@ -74,7 +74,7 @@ std::optional<uint64_t> WireDecoder3::SimpleValueReader::Read64(
 std::optional<double> WireDecoder3::SimpleValueReader::ReadDouble(
     std::span<const uint8_t>* in) {
   if (auto val = Read64(in)) {
-    return wpi::bit_cast<double>(val.value());
+    return std::bit_cast<double>(val.value());
   } else {
     return std::nullopt;
   }
diff --git a/ntcore/src/main/native/cpp/net3/WireEncoder3.cpp b/ntcore/src/main/native/cpp/net3/WireEncoder3.cpp
index 7f587c9f604..0ac2bd38ec2 100644
--- a/ntcore/src/main/native/cpp/net3/WireEncoder3.cpp
+++ b/ntcore/src/main/native/cpp/net3/WireEncoder3.cpp
@@ -4,8 +4,9 @@
 
 #include "WireEncoder3.h"
 
+#include <bit>
+
 #include <wpi/Endian.h>
-#include <wpi/MathExtras.h>
 #include <wpi/SmallVector.h>
 #include <wpi/leb128.h>
 #include <wpi/raw_ostream.h>
@@ -33,7 +34,7 @@ static void Write32(wpi::raw_ostream& os, uint32_t val) {
 
 static void WriteDouble(wpi::raw_ostream& os, double val) {
   uint8_t buf[8];
-  wpi::support::endian::write64be(buf, wpi::bit_cast<uint64_t>(val));
+  wpi::support::endian::write64be(buf, std::bit_cast<uint64_t>(val));
   os << buf;
 }
 
diff --git a/simulation/halsim_xrp/src/main/native/cpp/XRP.cpp b/simulation/halsim_xrp/src/main/native/cpp/XRP.cpp
index ffa1a5e56c1..812ebacb9bd 100644
--- a/simulation/halsim_xrp/src/main/native/cpp/XRP.cpp
+++ b/simulation/halsim_xrp/src/main/native/cpp/XRP.cpp
@@ -4,11 +4,11 @@
 
 #include "XRP.h"
 
+#include <bit>
 #include <string>
 
 #include <fmt/format.h>
 #include <wpi/Endian.h>
-#include <wpi/MathExtras.h>
 #include <wpi/json.h>
 
 using namespace wpilibxrp;
@@ -245,7 +245,7 @@ void XRP::SetupMotorTag(wpi::raw_uv_ostream& buf) {
 
     // Convert the value
     wpi::support::endian::write32be(value,
-                                    wpi::bit_cast<uint32_t>(motor.second));
+                                    std::bit_cast<uint32_t>(motor.second));
     buf << value[0] << value[1] << value[2] << value[3];
   }
 }
@@ -261,7 +261,7 @@ void XRP::SetupServoTag(wpi::raw_uv_ostream& buf) {
 
     // Convert the value
     wpi::support::endian::write32be(value,
-                                    wpi::bit_cast<uint32_t>(servo.second));
+                                    std::bit_cast<uint32_t>(servo.second));
     buf << value[0] << value[1] << value[2] << value[3];
   }
 }
@@ -283,17 +283,17 @@ void XRP::ReadGyroTag(std::span<const uint8_t> packet) {
 
   packet = packet.subspan(2);  // Skip past the size and tag
   float rate_x =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
   float rate_y =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[4]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[4]));
   float rate_z =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[8]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[8]));
   float angle_x =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[12]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[12]));
   float angle_y =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[16]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[16]));
   float angle_z =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[20]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[20]));
 
   // Make the json object
   wpi::json gyroJson;
@@ -314,11 +314,11 @@ void XRP::ReadAccelTag(std::span<const uint8_t> packet) {
 
   packet = packet.subspan(2);  // Skip past the size and tag
   float accel_x =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
   float accel_y =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[4]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[4]));
   float accel_z =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[8]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[8]));
 
   wpi::json accelJson;
   accelJson["type"] = "Accel";
@@ -409,7 +409,7 @@ void XRP::ReadAnalogTag(std::span<const uint8_t> packet) {
 
   packet = packet.subspan(3);
   float voltage =
-      wpi::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
+      std::bit_cast<float>(wpi::support::endian::read32be(&packet[0]));
 
   wpi::json analogJson;
   analogJson["type"] = "AI";
diff --git a/wpinet/src/netconsoleServer/native/cpp/main.cpp b/wpinet/src/netconsoleServer/native/cpp/main.cpp
index f65ada535a0..8c6a1626855 100644
--- a/wpinet/src/netconsoleServer/native/cpp/main.cpp
+++ b/wpinet/src/netconsoleServer/native/cpp/main.cpp
@@ -8,15 +8,14 @@
 #include <pty.h>
 #endif
 
+#include <bit>
 #include <cstdio>
 #include <memory>
 #include <string>
 
 #include <fmt/format.h>
-#include <wpi/MathExtras.h>
 #include <wpi/SmallVector.h>
 #include <wpi/StringExtras.h>
-#include <wpi/bit.h>
 #include <wpi/print.h>
 #include <wpi/timestamp.h>
 
@@ -52,7 +51,7 @@ static bool NewlineBuffer(std::string& rem, uv::Buffer& buf, size_t len,
   if (tcp) {
     // Header is 2 byte len, 1 byte type, 4 byte timestamp, 2 byte sequence num
     uint32_t ts =
-        wpi::bit_cast<uint32_t, float>((wpi::Now() - startTime) * 1.0e-6);
+        std::bit_cast<uint32_t, float>((wpi::Now() - startTime) * 1.0e-6);
     uint16_t len = rem.size() + toCopy.size() + 1 + 4 + 2;
     const uint8_t header[] = {static_cast<uint8_t>((len >> 8) & 0xff),
                               static_cast<uint8_t>(len & 0xff),
diff --git a/wpinet/src/netconsoleTee/native/cpp/main.cpp b/wpinet/src/netconsoleTee/native/cpp/main.cpp
index bd3da809e9a..142719ba75d 100644
--- a/wpinet/src/netconsoleTee/native/cpp/main.cpp
+++ b/wpinet/src/netconsoleTee/native/cpp/main.cpp
@@ -2,15 +2,14 @@
 // Open Source Software; you can modify and/or share it under the terms of
 // the WPILib BSD license file in the root directory of this project.
 
+#include <bit>
 #include <cstdio>
 #include <memory>
 #include <string>
 
 #include <fmt/format.h>
-#include <wpi/MathExtras.h>
 #include <wpi/SmallVector.h>
 #include <wpi/StringExtras.h>
-#include <wpi/bit.h>
 #include <wpi/print.h>
 #include <wpi/timestamp.h>
 
@@ -43,7 +42,7 @@ static bool NewlineBuffer(std::string& rem, uv::Buffer& buf, size_t len,
   if (tcp) {
     // Header is 2 byte len, 1 byte type, 4 byte timestamp, 2 byte sequence num
     uint32_t ts =
-        wpi::bit_cast<uint32_t, float>((wpi::Now() - startTime) * 1.0e-6);
+        std::bit_cast<uint32_t, float>((wpi::Now() - startTime) * 1.0e-6);
     uint16_t len = rem.size() + toCopy.size() + 1 + 4 + 2;
     const uint8_t header[] = {static_cast<uint8_t>((len >> 8) & 0xff),
                               static_cast<uint8_t>(len & 0xff),
diff --git a/wpiutil/src/main/native/cpp/DataLog.cpp b/wpiutil/src/main/native/cpp/DataLog.cpp
index 86649beecf4..705da5c0d37 100644
--- a/wpiutil/src/main/native/cpp/DataLog.cpp
+++ b/wpiutil/src/main/native/cpp/DataLog.cpp
@@ -377,7 +377,7 @@ void DataLog::AppendFloat(int entry, float value, int64_t timestamp) {
   if constexpr (std::endian::native == std::endian::little) {
     std::memcpy(buf, &value, 4);
   } else {
-    wpi::support::endian::write32le(buf, wpi::bit_cast<uint32_t>(value));
+    wpi::support::endian::write32le(buf, std::bit_cast<uint32_t>(value));
   }
 }
 
@@ -393,7 +393,7 @@ void DataLog::AppendDouble(int entry, double value, int64_t timestamp) {
   if constexpr (std::endian::native == std::endian::little) {
     std::memcpy(buf, &value, 8);
   } else {
-    wpi::support::endian::write64le(buf, wpi::bit_cast<uint64_t>(value));
+    wpi::support::endian::write64le(buf, std::bit_cast<uint64_t>(value));
   }
 }
 
@@ -508,14 +508,14 @@ void DataLog::AppendFloatArray(int entry, std::span<const float> arr,
     while ((arr.size() * 4) > kBlockSize) {
       buf = Reserve(kBlockSize);
       for (auto val : arr.subspan(0, kBlockSize / 4)) {
-        wpi::support::endian::write32le(buf, wpi::bit_cast<uint32_t>(val));
+        wpi::support::endian::write32le(buf, std::bit_cast<uint32_t>(val));
         buf += 4;
       }
       arr = arr.subspan(kBlockSize / 4);
     }
     buf = Reserve(arr.size() * 4);
     for (auto val : arr) {
-      wpi::support::endian::write32le(buf, wpi::bit_cast<uint32_t>(val));
+      wpi::support::endian::write32le(buf, std::bit_cast<uint32_t>(val));
       buf += 4;
     }
   }
@@ -540,14 +540,14 @@ void DataLog::AppendDoubleArray(int entry, std::span<const double> arr,
     while ((arr.size() * 8) > kBlockSize) {
       buf = Reserve(kBlockSize);
       for (auto val : arr.subspan(0, kBlockSize / 8)) {
-        wpi::support::endian::write64le(buf, wpi::bit_cast<uint64_t>(val));
+        wpi::support::endian::write64le(buf, std::bit_cast<uint64_t>(val));
         buf += 8;
       }
       arr = arr.subspan(kBlockSize / 8);
     }
     buf = Reserve(arr.size() * 8);
     for (auto val : arr) {
-      wpi::support::endian::write64le(buf, wpi::bit_cast<uint64_t>(val));
+      wpi::support::endian::write64le(buf, std::bit_cast<uint64_t>(val));
       buf += 8;
     }
   }
diff --git a/wpiutil/src/main/native/cpp/DataLogReader.cpp b/wpiutil/src/main/native/cpp/DataLogReader.cpp
index 17b2c9fa018..7eb88a6518b 100644
--- a/wpiutil/src/main/native/cpp/DataLogReader.cpp
+++ b/wpiutil/src/main/native/cpp/DataLogReader.cpp
@@ -4,11 +4,11 @@
 
 #include "wpi/DataLogReader.h"
 
+#include <bit>
 #include <utility>
 
 #include "wpi/DataLog.h"
 #include "wpi/Endian.h"
-#include "wpi/MathExtras.h"
 
 using namespace wpi::log;
 
@@ -97,7 +97,7 @@ bool DataLogRecord::GetFloat(float* value) const {
   if (m_data.size() != 4) {
     return false;
   }
-  *value = wpi::bit_cast<float>(wpi::support::endian::read32le(m_data.data()));
+  *value = std::bit_cast<float>(wpi::support::endian::read32le(m_data.data()));
   return true;
 }
 
@@ -105,7 +105,7 @@ bool DataLogRecord::GetDouble(double* value) const {
   if (m_data.size() != 8) {
     return false;
   }
-  *value = wpi::bit_cast<double>(wpi::support::endian::read64le(m_data.data()));
+  *value = std::bit_cast<double>(wpi::support::endian::read64le(m_data.data()));
   return true;
 }
 
@@ -143,7 +143,7 @@ bool DataLogRecord::GetFloatArray(std::vector<float>* arr) const {
   arr->reserve(m_data.size() / 4);
   for (size_t pos = 0; pos < m_data.size(); pos += 4) {
     arr->push_back(
-        wpi::bit_cast<float>(wpi::support::endian::read32le(&m_data[pos])));
+        std::bit_cast<float>(wpi::support::endian::read32le(&m_data[pos])));
   }
   return true;
 }
@@ -156,7 +156,7 @@ bool DataLogRecord::GetDoubleArray(std::vector<double>* arr) const {
   arr->reserve(m_data.size() / 8);
   for (size_t pos = 0; pos < m_data.size(); pos += 8) {
     arr->push_back(
-        wpi::bit_cast<double>(wpi::support::endian::read64le(&m_data[pos])));
+        std::bit_cast<double>(wpi::support::endian::read64le(&m_data[pos])));
   }
   return true;
 }