Sepa payments rejected because of exagerated decimal values

Recent regression on 6.8 (since september) with SEPA payments (CFONB flavour used)...

Generating a payment manually (not using 'pay' lines), the amount placed in the block is not a valid amount and is rejected by the bank.

Here is an extract of a sepa file generated:

    <CstmrCdtTrfInitn>
        <GrpHdr>
            <MsgId>c1a357050d75469da81c1901337f975b</MsgId>
            <CreDtTm>2023-09-21T09:23:26</CreDtTm>
            <NbOfTxs>1</NbOfTxs>
            <CtrlSum>435.1700000000000000</CtrlSum>
            <InitgPty>
--
        <PmtInf>
            <PmtInfId>d83b626990d74254853964b594e646d1</PmtInfId>
            <PmtMtd>TRF</PmtMtd>
            <BtchBookg>false</BtchBookg>
            <NbOfTxs>1</NbOfTxs>
            <CtrlSum>435.1700000000000000</CtrlSum>
--
            <Amt>
                <InstdAmt Ccy="EUR">435.1700000000000000</InstdAmt>
            </Amt>

Here the value should be '435.17' and not '435.1700000000000000'

Is it possible that some [python] library changed? In any event, perhaps there should be a secondary control made on the values just as there should be on the AlphaNumeric values to avoid illegal characters.